new_SCRNNs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "new_SCRNNs.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "FAETLXVZxYVq",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "b404656d-662d-430f-ad94-b22910ca669f"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from keras.layers import Input,Dense,Flatten,MaxPool2D,MaxPool1D,Activation,LeakyReLU,LSTM,BatchNormalization,Dropout,Conv2D,Conv1D,Lambda\n",
        "from keras.models import Model,load_model\n",
        "from keras.optimizers import Adam\n",
        "import keras.backend as K\n",
        "from keras.constraints import max_norm\n",
        "import h5py\n",
        "import keras\n",
        "import utils"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mEHmJDLuzRoP",
        "colab_type": "code",
        "outputId": "a5b9fe9d-8cf8-4d6f-9aae-5e3011ca1292",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive') "
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iEMxItb8x6GY",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "feat_file='/content/gdrive/My Drive/EE5611_project_features/new_scrnn_features.h5'\n",
        "model_dir = '/content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/'\n",
        "fid=h5py.File(feat_file,'r');\n",
        "x_tr = fid['x_tr']\n",
        "y_tr = fid['y_tr']\n",
        "x_ts=fid['x_ts']\n",
        "y_ts=fid['y_ts']\n",
        "snr_ts = np.array(fid['snr_ts'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "shyc3Orzxlru",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def new_scrnn():\n",
        "\n",
        "\t# Dropout and regularization not added. Make modifications to add those layers\n",
        "\t# change the hyperparameter values as mentioned in the cited paper\n",
        "\n",
        "\tinputs = Input((128,4,))\n",
        "\tl = BatchNormalization()(inputs)\n",
        "\t# l = Lambda(lambda t: K.expand_dims(t, -2))(l)\n",
        "\tl = Conv1D(filters=128,kernel_size=5,activation='relu')(l)\n",
        "\tl = MaxPool1D(3)(l)\n",
        "\tl = Conv1D(filters=128,kernel_size=5,activation='relu')(l)\n",
        "\t# l = Lambda(lambda t: K.squeeze(t, -2))(l)\n",
        "\tl = LSTM(128,return_sequences=True,activation='tanh',unroll=True)(l)\n",
        "\tl = Dropout(0.2)(l)\n",
        "\tl = LSTM(128,return_sequences=True,activation='tanh',unroll=True)(l)\n",
        "\tl = Dropout(0.2)(l)\n",
        "\tl = Flatten()(l)\n",
        "\toutputs = Dense(11,activation='softmax',kernel_constraint = max_norm(2.))(l)\n",
        "\n",
        "\tmodel = Model(inputs,outputs)\n",
        "\tmodel.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate=0.001) ,metrics=['accuracy'])\n",
        "\tmodel.summary()\n",
        "\treturn model\n",
        "\n",
        "def baseline_scrnn_tanh():\n",
        "\n",
        "\t# Dropout and regularization not added. Make modifications to add those layers\n",
        "\t# change the hyperparameter values as mentioned in the cited paper\n",
        "\n",
        "\tinputs = Input((128,2,))\n",
        "\tl = BatchNormalization()(inputs)\n",
        "\t# l = Lambda(lambda t: K.expand_dims(t, -2))(l)\n",
        "\tl = Conv1D(filters=128,kernel_size=5,activation='relu')(l)\n",
        "\tl = MaxPool1D(3)(l)\n",
        "\tl = Conv1D(filters=128,kernel_size=5,activation='relu')(l)\n",
        "\t# l = Lambda(lambda t: K.squeeze(t, -2))(l)\n",
        "\tl = LSTM(128,return_sequences=True,activation='tanh',unroll=True)(l)\n",
        "\tl = Dropout(0.2)(l)\n",
        "\tl = LSTM(128,return_sequences=True,activation='tanh',unroll=True)(l)\n",
        "\tl = Dropout(0.2)(l)\n",
        "\tl = Flatten()(l)\n",
        "\toutputs = Dense(11,activation='softmax',kernel_constraint = max_norm(2.))(l)\n",
        "\n",
        "\tmodel = Model(inputs,outputs)\n",
        "\tmodel.compile(loss='categorical_crossentropy', optimizer=Adam(learning_rate=0.001) ,metrics=['accuracy'])\n",
        "\tmodel.summary()\n",
        "\treturn model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HvoxSGxlx4II",
        "colab_type": "code",
        "outputId": "ba52d18e-dd1d-4d03-c0f8-8ebd9f269504",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 527
        }
      },
      "source": [
        "m_new_scrnn = new_scrnn()"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         (None, 128, 4)            0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_1 (Batch (None, 128, 4)            16        \n",
            "_________________________________________________________________\n",
            "conv1d_1 (Conv1D)            (None, 124, 128)          2688      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_1 (MaxPooling1 (None, 41, 128)           0         \n",
            "_________________________________________________________________\n",
            "conv1d_2 (Conv1D)            (None, 37, 128)           82048     \n",
            "_________________________________________________________________\n",
            "lstm_1 (LSTM)                (None, 37, 128)           131584    \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 37, 128)           0         \n",
            "_________________________________________________________________\n",
            "lstm_2 (LSTM)                (None, 37, 128)           131584    \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 37, 128)           0         \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 4736)              0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 11)                52107     \n",
            "=================================================================\n",
            "Total params: 400,027\n",
            "Trainable params: 400,019\n",
            "Non-trainable params: 8\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CNgLTGNw0H3a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "checkpoint = keras.callbacks.ModelCheckpoint(model_dir+ \"model_epoch-{epoch:02d}_val_acc-{val_accuracy:.2f}.h5\", monitor='val_accuracy', verbose=2, save_best_only=False, save_weights_only=False, mode='auto', period=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LioEJf3U0M5l",
        "colab_type": "code",
        "outputId": "f90003e8-c9f9-49ea-dd60-745a11d65d74",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history = m_new_scrnn.fit(x=x_tr,\n",
        "       y=y_tr,\n",
        "       validation_data=[x_ts,y_ts],\n",
        "       batch_size=400,\n",
        "       epochs=50,\n",
        "       shuffle='batch',\n",
        "       callbacks = [checkpoint],\n",
        "       verbose=1)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 198000 samples, validate on 22000 samples\n",
            "Epoch 1/50\n",
            "198000/198000 [==============================] - 45s 228us/step - loss: 1.5570 - accuracy: 0.4071 - val_loss: 1.8778 - val_accuracy: 0.3222\n",
            "\n",
            "Epoch 00001: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-01_val_acc-0.32.h5\n",
            "Epoch 2/50\n",
            "198000/198000 [==============================] - 26s 130us/step - loss: 1.2653 - accuracy: 0.5088 - val_loss: 1.2161 - val_accuracy: 0.5311\n",
            "\n",
            "Epoch 00002: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-02_val_acc-0.53.h5\n",
            "Epoch 3/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 1.2115 - accuracy: 0.5322 - val_loss: 1.1874 - val_accuracy: 0.5421\n",
            "\n",
            "Epoch 00003: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-03_val_acc-0.54.h5\n",
            "Epoch 4/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.1796 - accuracy: 0.5451 - val_loss: 1.1640 - val_accuracy: 0.5523\n",
            "\n",
            "Epoch 00004: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-04_val_acc-0.55.h5\n",
            "Epoch 5/50\n",
            "198000/198000 [==============================] - 25s 124us/step - loss: 1.1524 - accuracy: 0.5563 - val_loss: 1.1474 - val_accuracy: 0.5490\n",
            "\n",
            "Epoch 00005: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-05_val_acc-0.55.h5\n",
            "Epoch 6/50\n",
            "198000/198000 [==============================] - 26s 131us/step - loss: 1.1386 - accuracy: 0.5623 - val_loss: 1.1400 - val_accuracy: 0.5573\n",
            "\n",
            "Epoch 00006: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-06_val_acc-0.56.h5\n",
            "Epoch 7/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.1317 - accuracy: 0.5652 - val_loss: 1.1284 - val_accuracy: 0.5682\n",
            "\n",
            "Epoch 00007: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-07_val_acc-0.57.h5\n",
            "Epoch 8/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.1225 - accuracy: 0.5702 - val_loss: 1.1328 - val_accuracy: 0.5674\n",
            "\n",
            "Epoch 00008: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-08_val_acc-0.57.h5\n",
            "Epoch 9/50\n",
            "198000/198000 [==============================] - 24s 124us/step - loss: 1.1140 - accuracy: 0.5740 - val_loss: 1.1212 - val_accuracy: 0.5666\n",
            "\n",
            "Epoch 00009: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-09_val_acc-0.57.h5\n",
            "Epoch 10/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.1058 - accuracy: 0.5797 - val_loss: 1.1091 - val_accuracy: 0.5795\n",
            "\n",
            "Epoch 00010: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-10_val_acc-0.58.h5\n",
            "Epoch 11/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.0964 - accuracy: 0.5872 - val_loss: 1.1124 - val_accuracy: 0.5808\n",
            "\n",
            "Epoch 00011: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-11_val_acc-0.58.h5\n",
            "Epoch 12/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.0870 - accuracy: 0.5933 - val_loss: 1.0935 - val_accuracy: 0.5945\n",
            "\n",
            "Epoch 00012: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-12_val_acc-0.59.h5\n",
            "Epoch 13/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 1.0754 - accuracy: 0.6005 - val_loss: 1.0912 - val_accuracy: 0.5983\n",
            "\n",
            "Epoch 00013: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-13_val_acc-0.60.h5\n",
            "Epoch 14/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.0662 - accuracy: 0.6065 - val_loss: 1.0926 - val_accuracy: 0.5989\n",
            "\n",
            "Epoch 00014: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-14_val_acc-0.60.h5\n",
            "Epoch 15/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.0567 - accuracy: 0.6118 - val_loss: 1.0871 - val_accuracy: 0.6011\n",
            "\n",
            "Epoch 00015: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-15_val_acc-0.60.h5\n",
            "Epoch 16/50\n",
            "198000/198000 [==============================] - 26s 129us/step - loss: 1.0467 - accuracy: 0.6163 - val_loss: 1.0764 - val_accuracy: 0.6060\n",
            "\n",
            "Epoch 00016: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-16_val_acc-0.61.h5\n",
            "Epoch 17/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 1.0367 - accuracy: 0.6221 - val_loss: 1.0844 - val_accuracy: 0.6021\n",
            "\n",
            "Epoch 00017: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-17_val_acc-0.60.h5\n",
            "Epoch 18/50\n",
            "198000/198000 [==============================] - 25s 124us/step - loss: 1.0277 - accuracy: 0.6259 - val_loss: 1.0847 - val_accuracy: 0.6052\n",
            "\n",
            "Epoch 00018: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-18_val_acc-0.61.h5\n",
            "Epoch 19/50\n",
            "198000/198000 [==============================] - 24s 124us/step - loss: 1.0195 - accuracy: 0.6306 - val_loss: 1.0753 - val_accuracy: 0.6101\n",
            "\n",
            "Epoch 00019: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-19_val_acc-0.61.h5\n",
            "Epoch 20/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.0113 - accuracy: 0.6340 - val_loss: 1.0869 - val_accuracy: 0.6062\n",
            "\n",
            "Epoch 00020: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-20_val_acc-0.61.h5\n",
            "Epoch 21/50\n",
            "198000/198000 [==============================] - 25s 124us/step - loss: 0.9998 - accuracy: 0.6395 - val_loss: 1.0889 - val_accuracy: 0.6092\n",
            "\n",
            "Epoch 00021: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-21_val_acc-0.61.h5\n",
            "Epoch 22/50\n",
            "198000/198000 [==============================] - 26s 129us/step - loss: 0.9907 - accuracy: 0.6429 - val_loss: 1.0981 - val_accuracy: 0.6085\n",
            "\n",
            "Epoch 00022: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-22_val_acc-0.61.h5\n",
            "Epoch 23/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9800 - accuracy: 0.6472 - val_loss: 1.0987 - val_accuracy: 0.6074\n",
            "\n",
            "Epoch 00023: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-23_val_acc-0.61.h5\n",
            "Epoch 24/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9694 - accuracy: 0.6511 - val_loss: 1.1010 - val_accuracy: 0.6106\n",
            "\n",
            "Epoch 00024: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-24_val_acc-0.61.h5\n",
            "Epoch 25/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 0.9574 - accuracy: 0.6565 - val_loss: 1.1042 - val_accuracy: 0.6100\n",
            "\n",
            "Epoch 00025: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-25_val_acc-0.61.h5\n",
            "Epoch 26/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.9460 - accuracy: 0.6618 - val_loss: 1.1261 - val_accuracy: 0.6068\n",
            "\n",
            "Epoch 00026: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-26_val_acc-0.61.h5\n",
            "Epoch 27/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9369 - accuracy: 0.6646 - val_loss: 1.1273 - val_accuracy: 0.6019\n",
            "\n",
            "Epoch 00027: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-27_val_acc-0.60.h5\n",
            "Epoch 28/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.9248 - accuracy: 0.6693 - val_loss: 1.1447 - val_accuracy: 0.6053\n",
            "\n",
            "Epoch 00028: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-28_val_acc-0.61.h5\n",
            "Epoch 29/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.9123 - accuracy: 0.6747 - val_loss: 1.1672 - val_accuracy: 0.6047\n",
            "\n",
            "Epoch 00029: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-29_val_acc-0.60.h5\n",
            "Epoch 30/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.9022 - accuracy: 0.6776 - val_loss: 1.1572 - val_accuracy: 0.6045\n",
            "\n",
            "Epoch 00030: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-30_val_acc-0.60.h5\n",
            "Epoch 31/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 0.8903 - accuracy: 0.6828 - val_loss: 1.1654 - val_accuracy: 0.6058\n",
            "\n",
            "Epoch 00031: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-31_val_acc-0.61.h5\n",
            "Epoch 32/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.8790 - accuracy: 0.6868 - val_loss: 1.1903 - val_accuracy: 0.6020\n",
            "\n",
            "Epoch 00032: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-32_val_acc-0.60.h5\n",
            "Epoch 33/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.8666 - accuracy: 0.6909 - val_loss: 1.2020 - val_accuracy: 0.6007\n",
            "\n",
            "Epoch 00033: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-33_val_acc-0.60.h5\n",
            "Epoch 34/50\n",
            "198000/198000 [==============================] - 25s 128us/step - loss: 0.8558 - accuracy: 0.6954 - val_loss: 1.2289 - val_accuracy: 0.5988\n",
            "\n",
            "Epoch 00034: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-34_val_acc-0.60.h5\n",
            "Epoch 35/50\n",
            "198000/198000 [==============================] - 26s 132us/step - loss: 0.8452 - accuracy: 0.6996 - val_loss: 1.2294 - val_accuracy: 0.5964\n",
            "\n",
            "Epoch 00035: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-35_val_acc-0.60.h5\n",
            "Epoch 36/50\n",
            "198000/198000 [==============================] - 25s 128us/step - loss: 0.8336 - accuracy: 0.7033 - val_loss: 1.2514 - val_accuracy: 0.5937\n",
            "\n",
            "Epoch 00036: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-36_val_acc-0.59.h5\n",
            "Epoch 37/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 0.8224 - accuracy: 0.7075 - val_loss: 1.2500 - val_accuracy: 0.5975\n",
            "\n",
            "Epoch 00037: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-37_val_acc-0.60.h5\n",
            "Epoch 38/50\n",
            "198000/198000 [==============================] - 25s 124us/step - loss: 0.8114 - accuracy: 0.7112 - val_loss: 1.2665 - val_accuracy: 0.5945\n",
            "\n",
            "Epoch 00038: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-38_val_acc-0.59.h5\n",
            "Epoch 39/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 0.8005 - accuracy: 0.7154 - val_loss: 1.2824 - val_accuracy: 0.5950\n",
            "\n",
            "Epoch 00039: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-39_val_acc-0.60.h5\n",
            "Epoch 40/50\n",
            "198000/198000 [==============================] - 25s 128us/step - loss: 0.7918 - accuracy: 0.7185 - val_loss: 1.2973 - val_accuracy: 0.5964\n",
            "\n",
            "Epoch 00040: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-40_val_acc-0.60.h5\n",
            "Epoch 41/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.7814 - accuracy: 0.7218 - val_loss: 1.3123 - val_accuracy: 0.5894\n",
            "\n",
            "Epoch 00041: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-41_val_acc-0.59.h5\n",
            "Epoch 42/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 0.7706 - accuracy: 0.7251 - val_loss: 1.3103 - val_accuracy: 0.5943\n",
            "\n",
            "Epoch 00042: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-42_val_acc-0.59.h5\n",
            "Epoch 43/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.7608 - accuracy: 0.7286 - val_loss: 1.3254 - val_accuracy: 0.5921\n",
            "\n",
            "Epoch 00043: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-43_val_acc-0.59.h5\n",
            "Epoch 44/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.7528 - accuracy: 0.7322 - val_loss: 1.3323 - val_accuracy: 0.5932\n",
            "\n",
            "Epoch 00044: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-44_val_acc-0.59.h5\n",
            "Epoch 45/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.7412 - accuracy: 0.7362 - val_loss: 1.3373 - val_accuracy: 0.5923\n",
            "\n",
            "Epoch 00045: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-45_val_acc-0.59.h5\n",
            "Epoch 46/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 0.7317 - accuracy: 0.7390 - val_loss: 1.3845 - val_accuracy: 0.5887\n",
            "\n",
            "Epoch 00046: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-46_val_acc-0.59.h5\n",
            "Epoch 47/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.7214 - accuracy: 0.7430 - val_loss: 1.3819 - val_accuracy: 0.5873\n",
            "\n",
            "Epoch 00047: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-47_val_acc-0.59.h5\n",
            "Epoch 48/50\n",
            "198000/198000 [==============================] - 24s 120us/step - loss: 0.7149 - accuracy: 0.7453 - val_loss: 1.3767 - val_accuracy: 0.5900\n",
            "\n",
            "Epoch 00048: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-48_val_acc-0.59.h5\n",
            "Epoch 49/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 0.7034 - accuracy: 0.7486 - val_loss: 1.4045 - val_accuracy: 0.5882\n",
            "\n",
            "Epoch 00049: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-49_val_acc-0.59.h5\n",
            "Epoch 50/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 0.6953 - accuracy: 0.7521 - val_loss: 1.4149 - val_accuracy: 0.5859\n",
            "\n",
            "Epoch 00050: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/model_epoch-50_val_acc-0.59.h5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5GF8j_8q0fbB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 769
        },
        "outputId": "2003657d-fdb9-456d-a49c-8ed549c8ca18"
      },
      "source": [
        "acc_new_scrnn = history.history['accuracy']\n",
        "val_acc_new_scrnn = history.history['val_accuracy']\n",
        "loss_new_scrnn = history.history['loss']\n",
        "val_loss_new_scrnn = history.history['val_loss']\n",
        "epochs = range(1, len(acc_new_scrnn) + 1)\n",
        "\n",
        "print('===================================================')\n",
        "print('new scrnn\\n\\n')\n",
        "print('acc_new_scrnn\\n',acc_new_scrnn)\n",
        "print('val_acc_new_scrnn\\n',val_acc_new_scrnn)\n",
        "print('loss_new_scrnn\\n',loss_new_scrnn)\n",
        "print('val_loss_new_scrnn\\n',val_loss_new_scrnn)\n",
        "\n",
        "\n",
        "plt.plot(epochs, acc_new_scrnn, 'bo', label='Training acc')\n",
        "plt.plot(epochs, val_acc_new_scrnn, 'b', label='Validation acc')\n",
        "plt.title('Training and validation accuracy for New SCRNN')\n",
        "plt.legend()\n",
        "plt.figure()\n",
        "plt.plot(epochs, loss_new_scrnn, 'bo', label='Training loss')\n",
        "plt.plot(epochs, val_loss_new_scrnn, 'b', label='Validation loss')\n",
        "plt.title('Training and validation loss for New Scrnn')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "===================================================\n",
            "new scrnn\n",
            "\n",
            "\n",
            "acc_new_scrnn\n",
            " [0.40706566, 0.5087778, 0.5321768, 0.5450707, 0.5563081, 0.56231314, 0.56522727, 0.57018185, 0.5740101, 0.5797222, 0.58723736, 0.59329295, 0.60046464, 0.6064899, 0.611798, 0.61629796, 0.62211615, 0.62591916, 0.6306313, 0.63396466, 0.63954544, 0.6428788, 0.6471818, 0.651101, 0.6565202, 0.6618232, 0.6645505, 0.6693434, 0.6746717, 0.67758584, 0.6827828, 0.6867879, 0.6908838, 0.69538385, 0.6996414, 0.70328283, 0.70745456, 0.7111616, 0.71542424, 0.71853536, 0.7217626, 0.7251111, 0.7286313, 0.732197, 0.7361515, 0.73895454, 0.7430101, 0.74530303, 0.7485909, 0.7520808]\n",
            "val_acc_new_scrnn\n",
            " [0.32222726941108704, 0.5311363339424133, 0.542090892791748, 0.5523181557655334, 0.5490454435348511, 0.5573182106018066, 0.5681818127632141, 0.5673636198043823, 0.5665909051895142, 0.5795000195503235, 0.5807726979255676, 0.5944545269012451, 0.5982727408409119, 0.5988636612892151, 0.6010909080505371, 0.6060000061988831, 0.6020908951759338, 0.6052272915840149, 0.6100909113883972, 0.606181800365448, 0.6092272996902466, 0.6085454821586609, 0.6073636412620544, 0.610590934753418, 0.6099545359611511, 0.6068181991577148, 0.6019091010093689, 0.6053181886672974, 0.6046817898750305, 0.6045454740524292, 0.6058181524276733, 0.602045476436615, 0.6006818413734436, 0.5988181829452515, 0.5964090824127197, 0.593681812286377, 0.5975000262260437, 0.5945454835891724, 0.5950454473495483, 0.5964090824127197, 0.5893636345863342, 0.5943182110786438, 0.592090904712677, 0.5931817889213562, 0.5922726988792419, 0.5887272953987122, 0.5873181819915771, 0.5900454521179199, 0.5881817936897278, 0.5859090685844421]\n",
            "loss_new_scrnn\n",
            " [1.5570329605930984, 1.2652881379079337, 1.2114568105851762, 1.1795545525021023, 1.1523549992628772, 1.1385529630111926, 1.1316783925499578, 1.1224599843073373, 1.113991258120296, 1.1058481993097247, 1.0963743329048157, 1.0870401767769244, 1.075433161764434, 1.0661914346194026, 1.0567067931396792, 1.0467412984732425, 1.0367029422461385, 1.0277015157420226, 1.0194726956011069, 1.0112855500645108, 0.9997884766019957, 0.9907061634641705, 0.9799967462366278, 0.9693941802689523, 0.9574037091900603, 0.9460305756992764, 0.9369162242821972, 0.9248461631813435, 0.9123281564375367, 0.9022240695327219, 0.8902547172825745, 0.8789762204343622, 0.8666100893357788, 0.8557578029054584, 0.8452396373556118, 0.8335625830322805, 0.822380984311152, 0.8114464833278848, 0.8005172943828082, 0.7917909653499873, 0.7813755030583853, 0.7705686189911582, 0.760810895760854, 0.752818881742882, 0.7411640628419741, 0.7316508883177631, 0.7214150454058792, 0.714942314046802, 0.7034291963384609, 0.6953385489155548]\n",
            "val_loss_new_scrnn\n",
            " [1.8777873499826951, 1.2161219263618643, 1.1873668509450825, 1.1639752093702556, 1.1474067789248446, 1.1399705278348515, 1.1283775642344898, 1.1327502271075818, 1.1211642987094819, 1.109087920980528, 1.1124437882564961, 1.0935389238155702, 1.091170734717426, 1.092588651010936, 1.0870967728601748, 1.0763964805244044, 1.084376609968868, 1.0846626419574021, 1.0752997157608413, 1.0868766939090246, 1.0889249904652718, 1.0980583413207734, 1.0986751542024484, 1.1009599357039075, 1.1042385434546635, 1.1260557878312698, 1.12727476656384, 1.1446545907576695, 1.1672284908338704, 1.1572476606784303, 1.1653504707892848, 1.190330752521318, 1.2020431998397478, 1.2289406721226193, 1.229374808758439, 1.251359751912506, 1.250030689390207, 1.2664933600198394, 1.282410091485574, 1.2972756560087542, 1.3123495852490041, 1.3102579786881423, 1.3253788676874882, 1.3322983471237504, 1.3372652645159344, 1.3845378271536901, 1.3819018958559768, 1.3767341176280752, 1.4045154734620486, 1.4149152457121421]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H7_iXDi1f-1P",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "7d0d2a7b-52c4-4499-d742-03a9c7a219db"
      },
      "source": [
        "# new scrnn model evaluation\n",
        "new_scrnn_model = utils.evaluate_model('/content/gdrive/My Drive/EE5611_project_features/SCRNN_models/new_scrnn_models/','model_epoch-24_val_acc-0.61.h5',np.array(x_ts),np.array(y_ts),snr_ts)\n",
        "print(new_scrnn_model)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "===============================================================\n",
            "\n",
            "SNR: -20dB, Validation Accuracy: 9.545454545454545\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.10      0.02      0.03       100\n",
            "           1       0.04      0.01      0.02       100\n",
            "           2       0.00      0.00      0.00       100\n",
            "           3       0.14      0.04      0.06       100\n",
            "           4       0.00      0.00      0.00       100\n",
            "           5       0.00      0.00      0.00       100\n",
            "           6       0.10      0.90      0.17       100\n",
            "           7       0.23      0.03      0.05       100\n",
            "           8       0.08      0.01      0.02       100\n",
            "           9       0.22      0.04      0.07       100\n",
            "          10       0.00      0.00      0.00       100\n",
            "\n",
            "    accuracy                           0.10      1100\n",
            "   macro avg       0.08      0.10      0.04      1100\n",
            "weighted avg       0.08      0.10      0.04      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 2  2  0  2  3  0 87  0  1  1  2]\n",
            " [ 7  1  1  0  2  1 81  1  3  0  3]\n",
            " [ 0  1  0  1  0  0 92  2  1  1  2]\n",
            " [ 3  2  0  4  1  0 83  2  1  4  0]\n",
            " [ 1  2  1  1  0  0 89  2  2  0  2]\n",
            " [ 5  3  2  6  3  0 78  0  1  1  1]\n",
            " [ 0  1  2  2  2  0 90  1  0  2  0]\n",
            " [ 1  4  2  2  1  1 83  3  0  1  2]\n",
            " [ 0  2  1  4  3  1 86  0  1  2  0]\n",
            " [ 1  1  0  2  1  0 87  2  0  4  2]\n",
            " [ 0  5  0  4  3  0 84  0  2  2  0]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -18dB, Validation Accuracy: 10.363636363636363\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.05      0.01      0.02       100\n",
            "           1       0.10      0.02      0.03       100\n",
            "           2       0.12      0.02      0.03       100\n",
            "           3       0.29      0.09      0.14       100\n",
            "           4       0.07      0.02      0.03       100\n",
            "           5       0.75      0.03      0.06       100\n",
            "           6       0.10      0.88      0.17       100\n",
            "           7       0.00      0.00      0.00       100\n",
            "           8       0.00      0.00      0.00       100\n",
            "           9       0.14      0.03      0.05       100\n",
            "          10       0.25      0.04      0.07       100\n",
            "\n",
            "    accuracy                           0.10      1100\n",
            "   macro avg       0.17      0.10      0.05      1100\n",
            "weighted avg       0.17      0.10      0.05      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 1  2  3  3  2  0 88  0  0  0  1]\n",
            " [ 1  2  0  4  2  0 87  0  1  2  1]\n",
            " [ 4  1  2  4  1  0 81  0  2  3  2]\n",
            " [ 1  1  0  9  2  0 81  0  1  4  1]\n",
            " [ 1  2  1  0  2  0 91  0  1  1  1]\n",
            " [ 3  8  1  1  7  3 69  4  1  1  2]\n",
            " [ 1  1  1  1  2  1 88  1  1  2  1]\n",
            " [ 3  1  2  4  4  0 82  0  2  2  0]\n",
            " [ 1  1  2  0  1  0 88  3  0  2  2]\n",
            " [ 2  1  1  5  4  0 80  2  1  3  1]\n",
            " [ 1  0  3  0  1  0 88  1  0  2  4]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -16dB, Validation Accuracy: 10.181818181818182\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.16      0.04      0.06       100\n",
            "           1       0.00      0.00      0.00       100\n",
            "           2       0.15      0.02      0.04       100\n",
            "           3       0.15      0.05      0.07       100\n",
            "           4       0.03      0.01      0.02       100\n",
            "           5       0.80      0.08      0.15       100\n",
            "           6       0.10      0.88      0.18       100\n",
            "           7       0.06      0.01      0.02       100\n",
            "           8       0.12      0.02      0.03       100\n",
            "           9       0.09      0.01      0.02       100\n",
            "          10       0.00      0.00      0.00       100\n",
            "\n",
            "    accuracy                           0.10      1100\n",
            "   macro avg       0.15      0.10      0.05      1100\n",
            "weighted avg       0.15      0.10      0.05      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 4  2  2  3  4  0 78  2  1  0  4]\n",
            " [ 5  0  1  0  0  0 91  0  0  1  2]\n",
            " [ 3  0  2  3  2  0 83  0  3  2  2]\n",
            " [ 2  1  0  5  3  0 83  2  1  3  0]\n",
            " [ 0  1  1  7  1  0 85  0  2  0  3]\n",
            " [ 6 15  3  1  7  8 42  5  7  0  6]\n",
            " [ 1  2  1  3  1  1 88  1  0  1  1]\n",
            " [ 1  1  0  1  3  0 93  1  0  0  0]\n",
            " [ 1  2  1  0  2  1 90  0  2  1  0]\n",
            " [ 0  0  1  8  3  0 83  3  1  1  0]\n",
            " [ 2  1  1  3  3  0 85  3  0  2  0]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -14dB, Validation Accuracy: 11.545454545454545\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.00      0.00      0.00       100\n",
            "           1       0.02      0.01      0.01       100\n",
            "           2       0.12      0.03      0.05       100\n",
            "           3       0.17      0.05      0.08       100\n",
            "           4       0.08      0.02      0.03       100\n",
            "           5       0.81      0.17      0.28       100\n",
            "           6       0.10      0.86      0.19       100\n",
            "           7       0.07      0.02      0.03       100\n",
            "           8       0.06      0.02      0.03       100\n",
            "           9       0.21      0.05      0.08       100\n",
            "          10       0.15      0.04      0.06       100\n",
            "\n",
            "    accuracy                           0.12      1100\n",
            "   macro avg       0.16      0.12      0.08      1100\n",
            "weighted avg       0.16      0.12      0.08      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 0  1  2  1  2  0 88  1  0  3  2]\n",
            " [ 1  1  0  1  3  0 91  0  1  0  2]\n",
            " [ 6  1  3  3  1  1 76  1  3  4  1]\n",
            " [ 1  3  0  5  0  0 82  1  0  7  1]\n",
            " [ 1  3  1  3  2  0 86  1  1  0  2]\n",
            " [ 2 26  4  1  5 17  5 11 22  0  7]\n",
            " [ 0  1  1  3  1  1 86  3  2  1  1]\n",
            " [ 0  3  3  1  5  1 85  2  0  0  0]\n",
            " [ 6  8  4  5  1  0 60  6  2  2  6]\n",
            " [ 1  3  1  4  4  0 81  1  0  5  0]\n",
            " [ 1  1  5  2  1  1 82  1  0  2  4]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -12dB, Validation Accuracy: 16.181818181818183\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.03      0.01      0.02       100\n",
            "           1       0.04      0.02      0.03       100\n",
            "           2       0.32      0.08      0.13       100\n",
            "           3       0.36      0.20      0.26       100\n",
            "           4       0.23      0.07      0.11       100\n",
            "           5       0.61      0.33      0.43       100\n",
            "           6       0.12      0.86      0.21       100\n",
            "           7       0.03      0.01      0.02       100\n",
            "           8       0.16      0.09      0.12       100\n",
            "           9       0.32      0.11      0.16       100\n",
            "          10       0.00      0.00      0.00       100\n",
            "\n",
            "    accuracy                           0.16      1100\n",
            "   macro avg       0.20      0.16      0.13      1100\n",
            "weighted avg       0.20      0.16      0.13      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 1  2  3  2  2  0 86  1  1  0  2]\n",
            " [ 1  2  1  1  2  0 87  2  2  1  1]\n",
            " [ 9  8  8  2  1  2 54  6  3  2  5]\n",
            " [ 3  0  2 20  1  0 54  2  1 11  6]\n",
            " [ 1  2  0  4  7  0 80  1  2  2  1]\n",
            " [ 0 17  2  0  1 33  1  7 36  0  3]\n",
            " [ 2  1  1  6  1  0 86  2  0  1  0]\n",
            " [ 3  1  3  1  5  0 85  1  0  1  0]\n",
            " [ 7 18  4  0  7 19 22  8  9  1  5]\n",
            " [ 2  1  1 16  2  0 65  1  0 11  1]\n",
            " [ 2  3  0  4  1  0 83  2  1  4  0]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -10dB, Validation Accuracy: 23.363636363636363\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.07      0.03      0.04       100\n",
            "           1       0.11      0.07      0.09       100\n",
            "           2       0.67      0.36      0.47       100\n",
            "           3       0.32      0.33      0.33       100\n",
            "           4       0.24      0.17      0.20       100\n",
            "           5       0.53      0.37      0.44       100\n",
            "           6       0.16      0.83      0.27       100\n",
            "           7       0.06      0.02      0.03       100\n",
            "           8       0.25      0.18      0.21       100\n",
            "           9       0.35      0.17      0.23       100\n",
            "          10       0.18      0.04      0.07       100\n",
            "\n",
            "    accuracy                           0.23      1100\n",
            "   macro avg       0.27      0.23      0.21      1100\n",
            "weighted avg       0.27      0.23      0.21      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 3  5  3  2  8  1 73  1  1  1  2]\n",
            " [ 5  7  1  8  3  3 69  0  2  0  2]\n",
            " [22  7 36  3  3  4 15  2  6  0  2]\n",
            " [ 0  0  0 33  9  0 35  1  1 21  0]\n",
            " [ 0  0  1 12 17  0 60  0  0  6  4]\n",
            " [ 1  8  1  0  0 37  1 11 41  0  0]\n",
            " [ 0  1  0  3  6  2 83  1  0  2  2]\n",
            " [ 3  3  1  5  4  3 76  2  2  1  0]\n",
            " [ 2 25  8  1  7 17  5 12 18  0  5]\n",
            " [ 2  2  0 34  8  0 34  2  0 17  1]\n",
            " [ 2  4  3  2  7  3 71  2  1  1  4]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -8dB, Validation Accuracy: 39.09090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.24      0.15      0.18       100\n",
            "           1       0.24      0.12      0.16       100\n",
            "           2       0.69      0.66      0.68       100\n",
            "           3       0.48      0.77      0.59       100\n",
            "           4       0.46      0.39      0.42       100\n",
            "           5       0.53      0.55      0.54       100\n",
            "           6       0.29      0.88      0.44       100\n",
            "           7       0.13      0.07      0.09       100\n",
            "           8       0.36      0.34      0.35       100\n",
            "           9       0.40      0.23      0.29       100\n",
            "          10       0.36      0.14      0.20       100\n",
            "\n",
            "    accuracy                           0.39      1100\n",
            "   macro avg       0.38      0.39      0.36      1100\n",
            "weighted avg       0.38      0.39      0.36      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[15  6 15  2  5  3 40  5  5  1  3]\n",
            " [ 4 12  1  3 11  4 39 14  5  2  5]\n",
            " [27  2 66  2  0  0  0  1  2  0  0]\n",
            " [ 0  0  0 77  2  0  3  0  1 16  1]\n",
            " [ 2  0  1  9 39  1 26  2  2 12  6]\n",
            " [ 0  0  1  0  0 55  0  6 38  0  0]\n",
            " [ 2  1  4  0  1  0 88  2  0  1  1]\n",
            " [ 7  7  3  4  6  4 49  7  4  1  8]\n",
            " [ 0 15  0  0  1 35  2 12 34  0  1]\n",
            " [ 0  3  0 60  9  0  5  0  0 23  0]\n",
            " [ 6  4  4  2 10  2 47  6  3  2 14]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -6dB, Validation Accuracy: 54.36363636363637\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.61      0.50      0.55       100\n",
            "           1       0.37      0.35      0.36       100\n",
            "           2       0.66      0.87      0.75       100\n",
            "           3       0.52      0.84      0.64       100\n",
            "           4       0.63      0.71      0.67       100\n",
            "           5       0.47      0.69      0.56       100\n",
            "           6       0.76      0.86      0.81       100\n",
            "           7       0.34      0.13      0.19       100\n",
            "           8       0.42      0.46      0.44       100\n",
            "           9       0.45      0.19      0.27       100\n",
            "          10       0.56      0.38      0.45       100\n",
            "\n",
            "    accuracy                           0.54      1100\n",
            "   macro avg       0.53      0.54      0.52      1100\n",
            "weighted avg       0.53      0.54      0.52      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[50 10 23  1  2  3  3  3  4  0  1]\n",
            " [ 6 35  6  1  2 13  4 11 11  0 11]\n",
            " [11  0 87  0  0  1  1  0  0  0  0]\n",
            " [ 0  0  0 84  2  0  0  0  0 14  0]\n",
            " [ 1  1  2  5 71  3  3  0  2  6  6]\n",
            " [ 0  0  1  0  0 69  1  1 28  0  0]\n",
            " [ 0  2  1  4  3  0 86  1  1  2  0]\n",
            " [10 25  3  0  6 12  9 13 11  0 11]\n",
            " [ 0  6  1  0  0 41  1  5 46  0  0]\n",
            " [ 0  0  0 68 11  0  0  0  1 19  1]\n",
            " [ 4 16  7  0 15  5  5  4  5  1 38]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -4dB, Validation Accuracy: 67.45454545454545\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.91      0.83      0.87       100\n",
            "           1       0.50      0.47      0.48       100\n",
            "           2       0.82      0.96      0.88       100\n",
            "           3       0.53      0.94      0.68       100\n",
            "           4       0.83      0.84      0.84       100\n",
            "           5       0.52      0.74      0.61       100\n",
            "           6       0.92      0.83      0.87       100\n",
            "           7       0.46      0.19      0.27       100\n",
            "           8       0.54      0.65      0.59       100\n",
            "           9       0.55      0.21      0.30       100\n",
            "          10       0.86      0.76      0.81       100\n",
            "\n",
            "    accuracy                           0.67      1100\n",
            "   macro avg       0.68      0.67      0.66      1100\n",
            "weighted avg       0.68      0.67      0.66      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[83  1 14  0  0  0  0  0  1  0  1]\n",
            " [ 0 47  1  0  0 15  2 17 14  0  4]\n",
            " [ 2  0 96  0  0  0  2  0  0  0  0]\n",
            " [ 0  0  0 94  0  0  0  0  0  6  0]\n",
            " [ 0  1  0  4 84  0  0  0  0 10  1]\n",
            " [ 0  0  0  0  0 74  0  0 26  0  0]\n",
            " [ 3  0  4  4  3  0 83  0  1  1  1]\n",
            " [ 3 35  2  1  1 21  2 19 11  0  5]\n",
            " [ 0  1  0  0  0 30  1  3 65  0  0]\n",
            " [ 0  0  0 72  7  0  0  0  0 21  0]\n",
            " [ 0  9  0  1  6  3  0  2  3  0 76]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -2dB, Validation Accuracy: 79.0\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      0.91      0.94       100\n",
            "           1       0.65      0.76      0.70       100\n",
            "           2       0.94      0.97      0.96       100\n",
            "           3       0.59      0.96      0.73       100\n",
            "           4       0.91      0.91      0.91       100\n",
            "           5       0.72      0.81      0.76       100\n",
            "           6       0.93      0.90      0.91       100\n",
            "           7       0.70      0.43      0.53       100\n",
            "           8       0.74      0.83      0.78       100\n",
            "           9       0.65      0.28      0.39       100\n",
            "          10       0.96      0.93      0.94       100\n",
            "\n",
            "    accuracy                           0.79      1100\n",
            "   macro avg       0.80      0.79      0.78      1100\n",
            "weighted avg       0.80      0.79      0.78      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[91  0  6  0  0  0  3  0  0  0  0]\n",
            " [ 1 76  0  0  1  2  1 13  5  0  1]\n",
            " [ 0  0 97  0  0  0  1  1  0  0  1]\n",
            " [ 0  0  0 96  0  0  0  0  0  4  0]\n",
            " [ 0  0  0  0 91  0  0  0  0  9  0]\n",
            " [ 0  0  0  0  0 81  1  2 16  0  0]\n",
            " [ 2  1  0  2  0  0 90  1  0  2  2]\n",
            " [ 0 37  0  0  0 12  1 43  7  0  0]\n",
            " [ 0  0  0  0  0 17  0  0 83  0  0]\n",
            " [ 0  0  0 65  7  0  0  0  0 28  0]\n",
            " [ 0  3  0  0  1  1  0  1  1  0 93]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 0dB, Validation Accuracy: 87.27272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00       100\n",
            "           1       0.94      0.95      0.95       100\n",
            "           2       1.00      1.00      1.00       100\n",
            "           3       0.60      0.91      0.72       100\n",
            "           4       0.93      0.99      0.96       100\n",
            "           5       0.75      0.83      0.79       100\n",
            "           6       0.99      0.92      0.95       100\n",
            "           7       0.96      0.86      0.91       100\n",
            "           8       0.79      0.77      0.78       100\n",
            "           9       0.76      0.38      0.51       100\n",
            "          10       1.00      0.99      0.99       100\n",
            "\n",
            "    accuracy                           0.87      1100\n",
            "   macro avg       0.88      0.87      0.87      1100\n",
            "weighted avg       0.88      0.87      0.87      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0  95   0   1   0   2   0   2   0   0   0]\n",
            " [  0   0 100   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0  91   0   0   0   0   0   9   0]\n",
            " [  0   0   0   0  99   0   0   0   0   1   0]\n",
            " [  0   0   0   0   0  83   1   0  16   0   0]\n",
            " [  1   0   0   3   2   0  92   0   0   2   0]\n",
            " [  0   5   0   0   0   5   0  86   4   0   0]\n",
            " [  0   0   0   0   0  21   0   2  77   0   0]\n",
            " [  0   0   0  57   5   0   0   0   0  38   0]\n",
            " [  0   1   0   0   0   0   0   0   0   0  99]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 2dB, Validation Accuracy: 90.36363636363636\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.99      0.99       100\n",
            "           1       0.99      1.00      1.00       100\n",
            "           2       1.00      0.99      0.99       100\n",
            "           3       0.62      0.96      0.76       100\n",
            "           4       1.00      0.99      0.99       100\n",
            "           5       0.84      0.86      0.85       100\n",
            "           6       0.97      0.93      0.95       100\n",
            "           7       0.98      0.94      0.96       100\n",
            "           8       0.84      0.84      0.84       100\n",
            "           9       0.85      0.44      0.58       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.90      1100\n",
            "   macro avg       0.92      0.90      0.90      1100\n",
            "weighted avg       0.92      0.90      0.90      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   0   0   0   0   1   0   0   0   0]\n",
            " [  0 100   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0  99   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0  96   0   0   0   0   0   4   0]\n",
            " [  0   0   0   0  99   0   0   0   0   1   0]\n",
            " [  0   0   0   0   0  86   0   1  13   0   0]\n",
            " [  2   0   0   2   0   0  93   0   0   3   0]\n",
            " [  0   1   0   0   0   1   1  94   3   0   0]\n",
            " [  0   0   0   0   0  15   0   1  84   0   0]\n",
            " [  0   0   0  56   0   0   0   0   0  44   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 4dB, Validation Accuracy: 90.27272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.99      0.99       100\n",
            "           1       0.96      0.98      0.97       100\n",
            "           2       0.98      0.99      0.99       100\n",
            "           3       0.62      1.00      0.77       100\n",
            "           4       0.98      1.00      0.99       100\n",
            "           5       0.87      0.84      0.85       100\n",
            "           6       0.93      0.88      0.90       100\n",
            "           7       0.98      0.96      0.97       100\n",
            "           8       0.84      0.87      0.86       100\n",
            "           9       1.00      0.42      0.59       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.90      1100\n",
            "   macro avg       0.92      0.90      0.90      1100\n",
            "weighted avg       0.92      0.90      0.90      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   0   0   0   0   1   0   0   0   0]\n",
            " [  0  98   0   0   0   0   2   0   0   0   0]\n",
            " [  1   0  99   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  84   2   0  14   0   0]\n",
            " [  0   3   2   2   2   1  88   1   1   0   0]\n",
            " [  0   1   0   0   0   0   2  96   1   0   0]\n",
            " [  0   0   0   0   0  12   0   1  87   0   0]\n",
            " [  0   0   0  58   0   0   0   0   0  42   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 6dB, Validation Accuracy: 88.36363636363636\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      1.00      0.99       100\n",
            "           1       0.97      0.97      0.97       100\n",
            "           2       0.98      0.97      0.97       100\n",
            "           3       0.58      0.99      0.73       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.87      0.81      0.84       100\n",
            "           6       0.91      0.80      0.85       100\n",
            "           7       0.96      0.96      0.96       100\n",
            "           8       0.82      0.89      0.85       100\n",
            "           9       0.97      0.33      0.49       100\n",
            "          10       0.98      1.00      0.99       100\n",
            "\n",
            "    accuracy                           0.88      1100\n",
            "   macro avg       0.91      0.88      0.88      1100\n",
            "weighted avg       0.91      0.88      0.88      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0  97   0   0   0   1   2   0   0   0   0]\n",
            " [  1   0  97   0   0   0   2   0   0   0   0]\n",
            " [  0   0   0  99   0   0   0   0   0   1   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  81   1   0  18   0   0]\n",
            " [  2   3   2   5   1   1  80   4   0   0   2]\n",
            " [  0   0   0   0   0   0   2  96   2   0   0]\n",
            " [  0   0   0   0   0  10   1   0  89   0   0]\n",
            " [  0   0   0  67   0   0   0   0   0  33   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 8dB, Validation Accuracy: 91.36363636363636\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      0.98      0.98       100\n",
            "           1       0.99      1.00      1.00       100\n",
            "           2       1.00      0.98      0.99       100\n",
            "           3       0.64      1.00      0.78       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.87      0.94      0.90       100\n",
            "           6       0.93      0.91      0.92       100\n",
            "           7       0.98      0.97      0.97       100\n",
            "           8       0.92      0.83      0.87       100\n",
            "           9       0.94      0.44      0.60       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.91      1100\n",
            "   macro avg       0.93      0.91      0.91      1100\n",
            "weighted avg       0.93      0.91      0.91      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 98   0   0   0   0   0   2   0   0   0   0]\n",
            " [  0 100   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0  98   0   0   0   2   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  94   0   0   6   0   0]\n",
            " [  3   0   0   2   0   0  91   1   0   3   0]\n",
            " [  0   1   0   0   0   1   0  97   1   0   0]\n",
            " [  0   0   0   0   0  13   3   1  83   0   0]\n",
            " [  0   0   0  55   1   0   0   0   0  44   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 10dB, Validation Accuracy: 90.9090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00       100\n",
            "           1       1.00      0.98      0.99       100\n",
            "           2       0.99      0.99      0.99       100\n",
            "           3       0.62      1.00      0.76       100\n",
            "           4       0.96      1.00      0.98       100\n",
            "           5       0.91      0.86      0.89       100\n",
            "           6       0.95      0.86      0.90       100\n",
            "           7       0.97      0.99      0.98       100\n",
            "           8       0.88      0.92      0.90       100\n",
            "           9       1.00      0.40      0.57       100\n",
            "          10       0.97      1.00      0.99       100\n",
            "\n",
            "    accuracy                           0.91      1100\n",
            "   macro avg       0.93      0.91      0.90      1100\n",
            "weighted avg       0.93      0.91      0.90      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0  98   0   0   0   0   2   0   0   0   0]\n",
            " [  0   0  99   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  86   0   1  13   0   0]\n",
            " [  1   0   1   2   4   1  86   2   0   0   3]\n",
            " [  0   0   0   0   0   0   1  99   0   0   0]\n",
            " [  0   0   0   0   0   7   1   0  92   0   0]\n",
            " [  0   0   0  60   0   0   0   0   0  40   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 12dB, Validation Accuracy: 89.9090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.98      0.97       100\n",
            "           1       0.99      0.98      0.98       100\n",
            "           2       1.00      0.98      0.99       100\n",
            "           3       0.60      1.00      0.75       100\n",
            "           4       0.98      0.99      0.99       100\n",
            "           5       0.89      0.85      0.87       100\n",
            "           6       0.94      0.83      0.88       100\n",
            "           7       0.98      0.98      0.98       100\n",
            "           8       0.85      0.91      0.88       100\n",
            "           9       0.91      0.39      0.55       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.90      1100\n",
            "   macro avg       0.92      0.90      0.89      1100\n",
            "weighted avg       0.92      0.90      0.89      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 98   0   0   0   0   0   2   0   0   0   0]\n",
            " [  0  98   0   0   0   1   1   0   0   0   0]\n",
            " [  2   0  98   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   1  99   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  85   0   0  15   0   0]\n",
            " [  2   1   0   4   2   0  83   2   1   4   1]\n",
            " [  0   0   0   0   0   0   2  98   0   0   0]\n",
            " [  0   0   0   0   0   9   0   0  91   0   0]\n",
            " [  0   0   0  61   0   0   0   0   0  39   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 14dB, Validation Accuracy: 89.72727272727273\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.97      0.97       100\n",
            "           1       0.98      0.98      0.98       100\n",
            "           2       0.98      0.98      0.98       100\n",
            "           3       0.61      1.00      0.76       100\n",
            "           4       0.96      1.00      0.98       100\n",
            "           5       0.85      0.89      0.87       100\n",
            "           6       0.89      0.90      0.90       100\n",
            "           7       0.99      0.98      0.98       100\n",
            "           8       0.89      0.80      0.84       100\n",
            "           9       1.00      0.37      0.54       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.90      1100\n",
            "   macro avg       0.92      0.90      0.89      1100\n",
            "weighted avg       0.92      0.90      0.89      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 97   0   0   0   0   0   3   0   0   0   0]\n",
            " [  0  98   0   0   0   0   2   0   0   0   0]\n",
            " [  1   0  98   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   1   0   0  89   0   0  10   0   0]\n",
            " [  1   2   1   1   4   0  90   0   0   0   1]\n",
            " [  0   0   0   0   0   0   2  98   0   0   0]\n",
            " [  0   0   0   0   0  16   3   1  80   0   0]\n",
            " [  0   0   0  63   0   0   0   0   0  37   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 16dB, Validation Accuracy: 91.0909090909091\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.99      0.99       100\n",
            "           1       0.97      0.98      0.98       100\n",
            "           2       0.99      0.99      0.99       100\n",
            "           3       0.63      1.00      0.78       100\n",
            "           4       0.97      1.00      0.99       100\n",
            "           5       0.91      0.85      0.88       100\n",
            "           6       0.97      0.87      0.92       100\n",
            "           7       0.98      0.98      0.98       100\n",
            "           8       0.85      0.92      0.88       100\n",
            "           9       0.98      0.44      0.61       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.91      1100\n",
            "   macro avg       0.93      0.91      0.91      1100\n",
            "weighted avg       0.93      0.91      0.91      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   1   0   0   0   0   0   0   0   0]\n",
            " [  0  98   0   0   0   0   2   0   0   0   0]\n",
            " [  1   0  99   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   1   0  85   0   0  14   0   0]\n",
            " [  1   3   0   1   3   0  87   1   2   1   1]\n",
            " [  0   0   0   0   0   2   0  98   0   0   0]\n",
            " [  0   0   0   0   0   6   1   1  92   0   0]\n",
            " [  0   0   0  56   0   0   0   0   0  44   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 18dB, Validation Accuracy: 90.81818181818181\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      0.96      0.96       100\n",
            "           1       0.99      0.99      0.99       100\n",
            "           2       0.98      0.98      0.98       100\n",
            "           3       0.63      1.00      0.78       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.88      0.86      0.87       100\n",
            "           6       0.94      0.90      0.92       100\n",
            "           7       1.00      0.99      0.99       100\n",
            "           8       0.87      0.88      0.88       100\n",
            "           9       0.96      0.43      0.59       100\n",
            "          10       0.97      1.00      0.99       100\n",
            "\n",
            "    accuracy                           0.91      1100\n",
            "   macro avg       0.93      0.91      0.90      1100\n",
            "weighted avg       0.93      0.91      0.90      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 96   0   2   0   0   0   2   0   0   0   0]\n",
            " [  0  99   0   0   0   0   1   0   0   0   0]\n",
            " [  1   0  98   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  86   1   0  13   0   0]\n",
            " [  2   0   0   1   1   1  90   0   0   2   3]\n",
            " [  0   0   0   0   0   0   1  99   0   0   0]\n",
            " [  0   1   0   0   0  11   0   0  88   0   0]\n",
            " [  0   0   0  57   0   0   0   0   0  43   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "[9.545454545454545, 10.363636363636363, 10.181818181818182, 11.545454545454545, 16.181818181818183, 23.363636363636363, 39.09090909090909, 54.36363636363637, 67.45454545454545, 79.0, 87.27272727272727, 90.36363636363636, 90.27272727272727, 88.36363636363636, 91.36363636363636, 90.9090909090909, 89.9090909090909, 89.72727272727273, 91.0909090909091, 90.81818181818181]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAAEWCAYAAAB/tMx4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nO3deXyU9dnv8c+VBRJISICEEAgSkB1kS2RRUQLaghXhsaC2LuBRqZbWtafSevqoXZ7a1ra2p4rFKqJFo1JR6wErYlhU9kV2ZUsgIexJSEhCkpnr/DF3YgwJWSbJzCTX+/WaV+beZr5zZ3LNL7/7nvsnqooxxpjAE+TrAMYYYxrGCrgxxgQoK+DGGBOgrIAbY0yAsgJujDEBygq4McYEKCvgxhgToKyAG6+IyEoRyRGRtr7O4s9E5CoR+VxE8kTkjIh8JiKXO8tmiYiKyE+rbJMpIuOd+0+KSKmIFIhIrvNYY33wUowfsQJuGkxEEoFxgAI3NvNzhzTn83lDRDoAHwD/F+gEdAeeAs5XWu0M8FMRibzIQ72pqhFADJAGvN00iU2gsAJuvHEnsA54BZhZeYGI9BCRd0TkpIicFpG/VVp2r4jsEZF8EdktIiOd+SoifSqt94qI/Nq5P95pkT4mIseABSLSUUQ+cJ4jx7mfUGn7TiKyQESOOsvfdebvFJEpldYLFZFTIjKi6gt0ct5QaTrEeb6RIhImIv90Xl+uiGwUkbhq9lM/AFV9Q1Vdqlqkqh+p6vZK6+wB1gKP1LbTVbUMWAR0F5FYJ1eM8/pznRb+GhGxv+8Wzn7Bxht34ikki4BvlxcvEQnG0+LMABLxtDhTnWUzgCedbTvgabmfruPzdcXTgu0JzMbz/l3gTF8CFAF/q7T+a0A7YDDQBfizM/9V4PZK610PZKvq1mqe8w3ge5Wmvw2cUtUteD60ooAeQGfgPidDVV8BLhFZKCKTRaRjDa/vF8BDItKphuUAiEgbPPvvNJDjzH4UyARigTjg53j+MzItmBVw0yAichWewvmWqm4GDgDfdxaPAroB/1tVz6lqsap+6iy7B/i9qm5Uj/2qmlHHp3UDT6jqeacVe1pV/6WqhaqaD/wGuMbJFw9MBu5T1RxVLVXVVc7j/BO43unaALgDT7GvzuvAjSLSzpn+Pp6iDlCKp3D3cVrWm1X1bNUHcOZdhaegvgicFJH3q7bWVXUbsBx4rIYsN4tILp4PiXuB6U5rvDxLPNDTea1r1C501OJZATcNNRP4SFVPOdOv83U3Sg8go1JxqawHnmLfECdVtbh8QkTaicjfRSRDRM4Cq4Fo5z+AHsAZVc2p+iCqehT4DPiuiETjKfSLqntCVd2Pp3tjilPEb3ReK3iK/n+AVKeb5vciElrD4+xR1VmqmgAMwfMB92w1q/43cH8NXTFvqWo0nhb2TiCp0rI/APuBj0TkoIjMrS6HaVkC5kCQ8R8iEg7cDAQ7/dEAbfEUz2HAEeASEQmppogfAS6t4aEL8XR5lOuKp1ugXNUW5aNAf2C0qh4TkeHAVkCc5+kkItGqmlvNcy3E899ACLBWVbNqfsUV3ShBwG6nqKOqpXgORj7lHNBdCnwJvHSRx0JV94rIK8APalj2DvD4RbY/JSKzgU0i8rqqZjv/gTwKPCoiQ4BPRGSjqq64WBYT2KwFbhpiGuACBgHDndtAYA2evtkNQDbwtIi0dw72Xels+w/gJyKSJB59RKSns2wb8H0RCRaRSTjdIRcRiac7IdfpN36ifIGqZgPLgOedg52hInJ1pW3fBUYCD+LpE7+YVOBbwP183fpGRFJE5DKnxX8WTzeGu+rGIjJARB4tP8AqIj3wfCCsq+H5ngLuAqJrCqSqX+Jp/f/UecwbnH0pQB6e388FWUzLYgXcNMRMYIGqHlbVY+U3PAcQb8PTAp4C9AEO42lF3wKgqm/j6at+HcjHU0jLD9o96GyX6zzOu7XkeBYIB07hKYYfVll+B56iuhc4ATxUvkBVi4B/Ab2Ady72JM6HwVrgCuDNSou6AovxFO89wCqq70vPB0YD60XknJN1J54Wc3XPd8h5nPYXy4Wn22S2iHQB+gIfAwVO1udVNa2W7U2AEzvOYVorEflvoJ+q3l7rysb4IesDN62S0+VyN55WujEBybpQTKsjIvfiOci5TFVX+zqPMQ1lXSjGGBOgrAVujDEBqln7wGNiYjQxMbFB2547d4727Ws7KO87ls87ls87ls87/p5v8+bNp1Q19oIFqtpst6SkJG2otLS0Bm/bHCyfdyyfdyyfd/w9H7BJq6mp1oVijDEBygq4McYEKCvgxhgToHz+RZ7S0lIyMzMpLi6+6HpRUVHs2bOnmVLVn+XzTtV8YWFhJCQkEBpa7cX9jDH4QQHPzMwkMjKSxMREPNfhqV5+fj6RkRcbbcq3LJ93KudTVU6fPk1mZia9evXycTJj/JfPu1CKi4vp3LnzRYu3aV1EhM6dO9f6X5kxrZ3PCzhgxdtcwN4TxtTO510oxhj/tvVwDp/tP0X7tiFEtwslKjyUqPA2RIWHVkyHBvtFW7DVqVMBF5EH8YzBJ8CLqvqsczW3N/EMWpsO3KzVDF9ljAk8Lrfy8Z7jvLj6IJsyav+zbt8m2FPY27UhKjyE6EoFvkN4KKcySwk7eJrese2JjWjr9/9h5RaWsD0zjx1ZeXxxJJdDp87RJiSI8NBgwtsEEx4aTLs25fdDKt33/GzXJpiw8nWceZfGRhAWGtyoOWst4M7wTPfiGai2BPhQRD7AMyr4ClV92hl/by41D8baYkRERFBQUMDRo0d54IEHWLx48QXrjB8/nmeeeYbk5OQaH+fZZ59l9uzZtGvnGUHs+uuv5/XXXyc6usZBWEyAUlXOnCshSISgICEkSAguvznz/EVRiYt/bcnkpU8PcejUORI6hvPElEF8NykBl0vJLSolr6iU3MIS8iruV/7pmX/gZIFnXlEpJWWegYEW7PIMQBTZNoRese3pHdOeXjER9I5tT+/Y9vSKaU+7Ns3fKVBwvoy9Z1zsW32QLzJz2ZGVR8bpworlvWLa07dLBC63UlTqIr+4jJP55ykscVFU6qKoxEVhSRnuWq4L+PEjV9OnS+OeSFCXvTUQWK+qhQAisgq4CZgKjHfWWQispBUU8HLdunWrtnjX1bPPPsvtt99eUcCXLl3aWNGaVVlZGSEh1hNXnTKXm/+3I5t5Kw+w91h+jeuJQLBUKurOLSRICBLPzw7hoYzrG8P4/l24PLETbUIat8viVMF5XlubwWvrMjhzroRhCVH87fsjmDS4KyGVukc6tm9Tr8dVVYpL3Xzw8Sq69LmMgycLOHTqHAdPnmNjeg7vbjv6jfXjo8IqinnvmAh6xbbn0pgIuncMJ7gRPuiKS13syT7L9sw8T7HOzGP/yQI8F2XdQ/focIYmRHHL5T0YlhDNkO5RRIXXfiqrqlLiclNc4qawtMwp6i6KS10VhT4+Ktzr/FXV5S9vJ/AbEemMZ/zB64FNQJx6hpoCOIZnpGyvPPXvXew+erbaZS6Xi+Dg+v/7MahbB56YMrjG5XPnzqVHjx7MmTMHgCeffJKQkBDS0tLIycmhtLSUX//610ydOvUb26Wnp3PDDTewc+dOioqKmDVrFrt372bAgAEUFRVVrHf//fezceNGioqKmD59Ok899RR//etfOXr0KCkpKcTExJCWlkZiYiKbNm0iJiaGP/3pT7z88ssA3HPPPTz00EOkp6czefJkrrrqKj7//HO6d+/Oe++9R3h49W+KF198kfnz51NSUkKfPn14/vnniYyM5Pjx49x3330cPHgQgHnz5nHFFVfw6quv8swzzyAiDB06lNdee41Zs2Zxww03MH36dODr/z5WrlzJL37xCzp27MjevXv56quvmDZtGkeOHKG4uJgHH3yQ2bNnA/Dhhx/y85//HJfLRUxMDMuXL6d///58/vnnxMbG4na76devH8uXL/fr0xzro6jExVubjvDimoNk5hTRt0sEP79+AG2Cg3ApuNxuXO4qP1Upcysul+JSxeX+5i07r5iFn2fw4ppDtG8TzJV9PMV8fP9YukU3vDAcOFnAP9Yc4p0tmZwvc3PtwC7cO643o3p1apRuDhEhvE0wse2CuKZfLNf0++b1mIpKXBw6dc4p6p7ifuDUOd7bdpT84rJKjwNtgoNoExJE25CgivsVt4rpYNoEO+tUml9S5mbn0Ty+PJZPmdNUjolow9CEaL4zNB7JOcxtk8cRE9G2wa+zbUgwbUOCiaL5vrtQp+uBi8jdwA+Bc8Au4DwwS1WjK62To6odq9l2Np7uFuLi4pJSU1O/sTwqKoo+ffoA8LuPDrD3eEG1GVS1QW+oAXERPPatmgZBhy+++IK5c+eybNkyAC6//HKWLFlChw4d6NChA6dPn2bChAls27YNESE+Pp7s7GwyMjK4+eabWb9+PX/729/YtWsX8+bNY+fOnYwbN44VK1YwcuRIzpw5Q6dOnXC5XEyZMoXf//73DBkyhCFDhrBq1So6d+4MUDF9+PBh7r//flasWIGqMmHCBF588UWio6MZPnw4q1atYujQocycOZPJkydz6623Vvu6Tp8+XfHYv/zlL4mNjeX+++9n1qxZXH755cyZMweXy1XRHfT973+fjz/+mM6dO1dkvu+++5g0aRLTpk0DqHjta9asYcaMGaxbt47yq0uWb1NUVMT48eNZunQpqsq4ceNYtmwZiYmJFev89re/pUOHDsyZM4cVK1awYMECFi5ceMEH9P79+8nLy6v377wpFBQUEBERcfF1SpQVh0v5OKOU/FLoEx3Ed3qHMiw2mKBGKIbFZcqeMy62n/TcThd7/nYTIoQBUW6Su4fTJzqIkFpaqqrKVzluPkwvZesJFyFBcFW3EL6VGEq3iKY5GFmX/Vc1Y34JZJ9zc6zQzekipdQNZe7ynzXfL1+v8n0RuCQyiF5RwSR2CKJXVBCdwqSiptQ3X3NLSUnZrKoX9MnW6X9fVX0JeAlARP4HzyC1x0UkXlWzRSQez6Cx1W07H5gPkJycrOPHj//G8j179lS0vH793eE1ZmiqL6JcddVVnD59mvz8fE6ePEnnzp3p06cPDz/8MKtXryYoKIjs7GwKCwvp2rUrAJGRkURERBAUFERkZCTr16/nnnvuITIykrFjxzJ06FDat29PZGQkixYtYv78+ZSVlVUU/rFjxyIiREREVLym8umtW7fy3e9+t+K5pk+fzpYtW7jxxhvp1asXV17pGdx99OjRHD9+vMZ9smXLFu644w5yc3MpKChgwoQJREZGsnr1al5//XXatvW0NKKjo1myZAm33HJLRTEuf8zQ0FDCw8O/8RyRkZG0a9eOUaNGcdlll1XM/+Mf/8iSJUsAyMrK4tixY5w8eZJrrrmmYr3yx7n//vuZOnUqc+fOJTU1lXvvvZfg4OALXktYWBgjRoxo4G+2ca1cuZKq791yR3OLeOnTQ7yx4TCFJS4mDujCfeMv5fLETtWu741Jzk9VZf+JAtK+PEHa3pOkHTrNx1nFRLYNYVy/GMb387TOu3QIq9i2zOXmP7uOM3/NQb44kkvHdqE8MLE3d47t2eCWZ11dbP/5A3/PV5O6noXSRVVPiMglePq/x+AZzXsm8LTz870mS9nEZsyYweLFizl27Bi33HILixYt4uTJk2zevJnQ0FASExMb9KWSQ4cO8cwzz7Bx40Y6duzIrFmzvPpySnnRBQgODv5GV01Vs2bN4t1332XYsGG88sorLF++vN7PFxISgtvtOQDldrspKSmpWFb52skrV67k448/Zu3atbRr147x48df9HX26NGDuLg4PvnkEzZs2MCiRYsoLCyscX1/te94Pi+sOsh727JQYOqwbvzgmkvp37Xpu4JEhL5xkfSNi2T21Zey7OM0pOtAVn3lKehLdxwDYHC3DozvH0t0eBteXZfOkTNFJHZux6+mDWH6yATC2zTuWRGmedX1/6V/ichu4N/AHFXNxVO4rxORfcC1znRAuuWWW0hNTWXx4sXMmDGDvLw8unTpQmhoKGlpaWRkZFx0+6uvvpq3334bgJ07d7J9+3YAzp49S/v27YmKiuL48eMV3TTgaY3m5194YGvcuHG8++67FBYWcu7cOZYsWcK4cePq/Zry8/OJj4+ntLSURYsWVcyfOHEi8+bNAzzHFfLy8pgwYQJvv/02p0+fBjzdIQCJiYls3rwZgPfff5/S0tJqnysvL4+OHTvSrl079u7dy7p1nrMNxowZw+rVqzl06NA3Hhc8ffu33347M2bMaNCxDV/anHGGexZu4ro/r2bpjmxuH9OTVf97PH+6ZXizFO/qhIcIk4Z05bc3DWXtzyaw7MFxPDZpAO3bhvDCqoP8Zuke4iLDeOH2JFY8Op47xvS04t0C1LUL5YIKoqqngYmNnsgHBg8eTH5+Pt27dyc+Pp7bbruNKVOmcNlll5GcnMyAAQMuuv3999/P7bffzsCBAxk4cCBJSUkADBs2jBEjRjBgwAB69OhR0f0BMHv2bCZNmkS3bt1IS0urmD9y5EhmzZrFqFGjAE+hGzFiBOnp6fV6Tb/61a8YPXo0sbGxjB49uqJ4/uUvf2H27Nm89NJLBAcHM2/ePMaOHcvjjz/ONddcQ3BwMCNGjOCVV17h3nvvZerUqQwbNoxJkybVOGLJpEmTeOGFFxg4cCD9+/dnzJgxAMTGxjJ//nxuuukm3G43Xbp0qfhP4MYbb+Suu+7irrvuqtfr8hVV5ZO9x3lh5UE2pJ8hul0oD07sy8wrEulUzzMzmpqIMDC+AwPjO3D/+EvJKyrlVMF5Lo313z5e0zDNOqhxcnKybtq06Rvz9uzZw8CBA2vdNpAuxuSP/C3fpk2bePjhh1mzZg1Qfb66vjea2tbDOfz41bVkFijdosK49+re3HJ5D5+cs1wTf+/DtXzeEZGGH8Q0pjE9/fTTzJs37xtdO/4qK7eIexZuAhf8ccYwbhzezb42bvyGvRMD3Jw5cxg+fPg3bgsWLPB1rIuaO3cuGRkZXHXVVb6OclHFpS7ue20z58vcPJocxneTEqx4G7/iFy3whp7jbeC5557zdYQm0ZxdezU9/8/f2cGOrDxevDOZ0BP+OxiGab183pwICwvj9OnTPv+DNf6jfECHsLCw2lduIgs+S+edrVk8fG0/rhvk9ZeMjWkSPm+BJyQkkJmZycmTJy+6XnFxsU//oGtj+bxTNV/5kGq+8PmBU/xm6R6+NSiOH0/o45MMxtSFzwt4aGhonYbNWrlypd98K686ls87/pIvM6eQH72+lcTO7fjjzcP86kqBxlTl8y4UY/xFUYmLH7y2mdIyN/PvTCYyzAZUNv7N5y1wY/yBqvKzd7azO/ss/7gz2b70YgKCtcCNAV769BDvbjvKI9f2Y+JAO2hpAoMVcNPqfb7/FL9dtpdvD45jToodtDSBwwq4adWOnClkzutb6B3Tnj/ePNwOWpqAYgXctFrlBy3L3Mr8O5OJaGuHhExgsXesaZVUlcf+tZ09x87y8szL6RVT/ZUWjfFn1gI3rdI/1hzi/S+O8pNv9SdlQBdfxzGmQayAm1bn032n+O2yPUwe0pUfjq95vFRj/F2dCriIPCwiu0Rkp4i8ISJhItJLRNaLyH4ReVNE/Ouq9sZU48iZQn70xhb6dIngmRnD7CJqJqDVWsBFpDvwAJCsqkOAYOBW4HfAn1W1D5AD3N2UQY3xVmFJGfe+ugm3W5l/RzLt7aClCXB17UIJAcJFJARoB2QDE4DFzvKFwLTGj2dM41BVfrp4O18ez+ev3xtBoh20NC1AnYZUE5EHgd8ARcBHwIPAOqf1jYj0AJY5LfSq284GZgPExcUlpaamNihoQUEBERH++/Vmy+edps639FAJb31ZyvR+odzQu/69fa19/3nL8nknJSWl2iHVUNWL3oCOwCdALBAKvAvcDuyvtE4PYGdtj5WUlKQNlZaW1uBtm4Pl805T5vts30ntNfcD/eE/N6vb7W7QY7Tm/dcYLJ93gE1aTU2tSxfKtcAhVT2pqqXAO8CVQLTTpQKQAGR58wljTFMoKXPz+Ls7Sezcnt9PH2oHLU2LUpcCfhgYIyLtxPPunwjsBtKA6c46M4H3miaiMQ336tp0Dp06xy+mDLKDlqbFqbWAq+p6PAcrtwA7nG3mA48Bj4jIfqAz8FIT5jSm3k4VnOcvK/Yxvn8sKf3tyzqm5alTk0RVnwCeqDL7IDCq0RMZ00j++NFXFJW4+D/fGeTrKMY0CfsmpmmRdh89y5sbD3PH2J706eK/ZxcY4w0r4KbFUVV++cEuosJDeWhiP1/HMabJWAE3Lc5/dh1j3cEzPHJdP6La2biWpuWyAm5alOJSF79Zuof+cZF8b9Qlvo5jTJOyAm5alJc/O8SRM0X895RBhATb29u0bPYONy3GibPFPPfJfq4bFMeVfWJ8HceYJmcF3LQYf/jPl5S43Dx+/UBfRzGmWVgBNy3Cjsw8Fm/J5H9d2cuuNGhaDSvgJuCpKk/9exed27fhRxP6+DqOMc3GCrgJeB9sz2ZTRg4/+VZ/IsPstEHTelgBNwGtuNTF08v2Mii+AzOSe/g6jjHNygq4CWjzVx8kK7eIJ6YMIjjILhVrWhcr4CZgHcsrZt7KA1x/WVdG9+7s6zjGNDsr4CZg/e7DvbhU+dlkO23QtE5WwE1A2nI4hyVbs7h3XC96dGrn6zjG+EStBVxE+ovItkq3syLykIh0EpHlIrLP+dmxOQIb43Yrv/z3brpEtuWH4+20QdN61WVEni9VdbiqDgeSgEJgCTAXWKGqfYEVzrQxTe69L7LYdiSXn04aYMOkmVatvl0oE4EDqpoBTAUWOvMXAtMaM5gx1SksKeN3y75kWEIUN43o7us4xviUeEasr+PKIi8DW1T1byKSq6rRznwBcsqnq2wzG5gNEBcXl5SamtqgoAUFBURE+O/IKpbPO3XN986+Et4/UMrjo8Po2zG4GZJ5tJT95yuWzzspKSmbVTX5ggWqWqcb0AY4BcQ507lVlufU9hhJSUnaUGlpaQ3etjlYPu/UJd+RM+e03+NL9cevb2n6QFW0hP3nS5bPO8Amraam1qcLZTKe1vdxZ/q4iMQDOD9PNPTTxZi6eHrZXkRg7uQBvo5ijF+oTwH/HvBGpen3gZnO/ZnAe40VypiqNhw6wwfbs/nB1ZfSLTrc13GM8Qt1KuAi0h64Dnin0uyngetEZB9wrTNtTKNzuz2DFMdHhXHfNZf6Oo4xfqNO52Cp6jmgc5V5p/GclWJMk1q8JZOdWWf5y63DCW/TfAcujfF39k1M49dKytz85eN9DEuI4sZh3Xwdxxi/YgXc+LW3Nx8hK7eIR77VH8/ZqsaYclbAjd86X+bib5/sZ+Ql0Vzd1wYpNqYqK+DGb7258QjZecU8cp21vo2pjhVw45eKS108l7afUYmduLKPXevbmOpYATd+6Y0Nhzl+9jwPXdfXWt/G1MAKuPE7xaUunl95gDG9O3HFpdb3bUxNrIAbv/PPdRmczD/Pw9f283UUY/yaFXDjVwpLynhh1QGu7NPZxrk0phZ2NXzjV15bm8GpghJesNa3MbWyFrjxG+fOl/H31Qe5ul8syYmdfB3HGL9nBdz4jYVr0zlzroSHr+3r6yjGBATrQjF+oahMmf/5QVL6xzLiEhsf25i6sBa48QvLM0rJLSzl4eus79uYurICbnzubHEpHx4q5dqBXRiacMGwqsaYGtR1QIdoEVksIntFZI+IjBWRTiKyXET2OT/t/17TIC9/eojCMnjIzjwxpl7q2gL/C/Chqg4AhgF7gLnAClXtC6xwpo2pl7zCUl769BBJccEM6R7l6zjGBJRaC7iIRAFXAy8BqGqJquYCU4GFzmoLgWlNFdK0XC99epD84jKm9Wnj6yjGBBzxjFh/kRVEhgPzgd14Wt+bgQeBLFWNdtYRIKd8usr2s4HZAHFxcUmpqakNClpQUEBERESDtm0Olq/+CkqUn6wqZEhMMLP6lvldvsr8cf9VZvm84+/5UlJSNqtq8gULVPWiNyAZKANGO9N/AX4F5FZZL6e2x0pKStKGSktLa/C2zcHy1d/vP9yjiXM/0L3ZZ/0yX2WWzzuWzzvAJq2mptalDzwTyFTV9c70YmAkcFxE4gGcnye8+YQxrcuZcyUs+Cyd71wWT/+ukb6OY0xAqrWAq+ox4IiI9HdmTcTTnfI+MNOZNxN4r0kSmhbp76sPUFTq4iH71qUxDVbXb2L+GFgkIm2Ag8BdeIr/WyJyN5AB3Nw0EU1Lc6rgPK9+nsHUYd3o08Va38Y0VJ0KuKpuw9MXXtXExo1jWoO/rzrA+TIXD0y01rcx3rBvYppmdSK/mNfWZTBtRHd6x/rvUX9jAoEVcNOsXlh5kFKX8sAEa30b4y0r4KbZHD9bzD/XZ3DTiO4kxrT3dRxjAp4VcNNsnk/bj9ut/Nha38Y0Civgpllk5xXxxoYjzEhO4JLO7Xwdx5gWwQq4aRbPpe1HUeak9PF1FGNaDCvgpsll5hTy5sYj3Jzcg4SO1vo2prFYATdN7u+rDiKItb6NaWRWwE2TyissZfHmTKYO70a36HBfxzGmRbECbppU6sbDFJW6uOvKXr6OYkyLYwXcNJkyl5tX12YwulcnBnXr4Os4xrQ4VsBNk1m++zhZuUXW+jamiVgBN01mwWfpJHQM57pBcb6OYkyLZAXcNImdWXlsSD/DrCsSCQ4SX8cxpkWyAm6axILP0mnXJpgZyT18HcWYFqtO1wMXkXQgH3ABZaqaLCKdgDeBRCAduFlVc5ompgkkJ/PP8+8vjnLrqB5EhYf6Oo4xLVZ9WuApqjpcvx4ZeS6wQlX7AiucaWN4ff1hSlxuZl6R6OsoxrRo3nShTAUWOvcXAtO8j2MC3fkyF6+ty2B8/1gutQEbjGlS4hmxvpaVRA4BOYACf1fV+SKSq6rRznIBcsqnq2w7G5gNEBcXl5SamtqgoAUFBURE+G9BsHwen2WV8uKOEh5NastlsXUdctX2n7csn3f8PV9KSsrmSr0fX1PVWm9Ad+dnF+AL4Gogt8o6ObU9TlJSkjZUWlpag7dtDpZP1e126w1/XaMTns9x2a0AABPQSURBVElTt9tdr21t/3nH8nnH3/MBm7SamlqnLhRVzXJ+ngCWAKOA4yISD+D8POHlh4wJcJszctiRlcesK3vh+afMGNOUai3gItJeRCLL7wPfAnYC7wMzndVmAu81VUgTGBZ8lk6HsBC+O7K7r6MY0yrUpZMyDljitKhCgNdV9UMR2Qi8JSJ3AxnAzU0X0/i7o7lFfLjrGPdc1Yt2bere922Mabha/9JU9SAwrJr5p4GJTRHKBJ5X12agqtwxtqevoxjTatg3MY3XikpcvLHhMN8e3NVG3DGmGVkBN157Z2smeUWldtVBY5qZFXDjFVXllc/SGdytA5cndvR1HGNaFSvgxiuf7j/FvhMF3GWnDhrT7KyAG68s+CydmIg2TBkW7+soxrQ6VsBNgx06dY5P9p7gttE9aRsS7Os4xrQ6VsBNgy38PJ3QYOG2MZf4OooxrZIVcNMgZ4tLeXvTEaYM7UaXyDBfxzGmVbICbhrkrY1HOFfislMHjfEhK+Cm3lxuZeHadJJ7duSyhChfxzGm1bICbuptxZ7jHDlTZK1vY3zMCriptwWfpdMtKoxvD47zdRRjWjUr4KZe9mSfZe3B09x5RSIhwfb2McaX7C/Q1Msrn6UTFhrErZf38HUUY1o9K+Cmzs6cK+HdbVncNDKB6HZtfB3HmFavzgVcRIJFZKuIfOBM9xKR9SKyX0TeFBH7i27h3thwmPNlbu66ItHXUYwx1K8F/iCwp9L074A/q2ofPCPW392YwYx/KXW5eXVtOuP6xtA3LtLXcYwx1LGAi0gC8B3gH860ABOAxc4qC4FpTRHQ+IelO7I5fvY8d12Z6OsoxhiHeEasr2UlkcXAb4FI4CfALGCd0/pGRHoAy1R1SDXbzgZmA8TFxSWlpqY2KGhBQQEREREN2rY5tPR8v1xbxLlS5bfjwglqgsvGtvT919Qsn3f8PV9KSspmVU2+YIGqXvQG3AA879wfD3wAxAD7K63TA9hZ22MlJSVpQ6WlpTV42+bQkvNtyTijPR/7QBd8erDxAlXRkvdfc7B83vH3fMAmraam1mX48CuBG0XkeiAM6AD8BYgWkRBVLQMSgCzvPmOMv3rl83Qi24YwPdlOHTTGn9TaB66qP1PVBFVNBG4FPlHV24A0YLqz2kzgvSZLaXzmRH4xS3dkMz05gYi2dfm8N8Y0F2/OA38MeERE9gOdgZcaJ5LxJ6kbjlDqUu4Y09PXUYwxVdSrSaWqK4GVzv2DwKjGj2T8RanLzaL1GVzdL5besf57gMeY1sq+iWlqtHz3cY6fPc/Msdb6NsYfWQE3NVr4eTo9OoUzvn8XX0cxxlTDCrip1t5jZ1l/6Ax3jOlJcFDjn/dtjPGeFXBTrVfXZtA2JIib7dRBY/yWFXBzgbyiUpZsyWLa8O521UFj/JgVcHOBxZszKSp1cYcdvDTGr1kBN9/gdiuvrU0nqWdHhnS3AYuN8WdWwM03rN53kvTThdxprW9j/J4VcPMNr67NICaiLZOHxPs6ijGmFlbATYWM0+dI+/IE3x99CW1C7K1hjL+zv1JT4Z/rMggW4bbRl/g6ijGmDqyAGwCKSly8tSmTbw/pSlyHMF/HMcbUgRVwA8D7X2SRV1TKzLGJvo5ijKkjK+AGVWXh5xkM6BrJ5YkdfR3HGFNHVsANmzNy2J19lplXJCJNMN6lMaZp1FrARSRMRDaIyBcisktEnnLm9xKR9SKyX0TeFBH7znWAWrg2gw5hIUwd3s3XUYwx9VCXFvh5YIKqDgOGA5NEZAzwO+DP6hmZPge4u+limqZy4mwxy3Zkc3NyD9q1sSHTjAkkdRkTU1W1wJkMdW4KTAAWO/MXAtOaJKFpUq9vOIxLldttyDRjAo54RqyvZSWRYGAz0Ad4DvgDsM5pfSMiPYBlqjqkmm1nA7MB4uLiklJTUxsUtKCggIgI/x3WKxDzlbmVR1cV0bNDEI8k+fbUwUDcf/7E8nnH3/OlpKRsVtXkCxaoap1vQDSe0eivAvZXmt8D2Fnb9klJSdpQaWlpDd62OQRivve3ZWnPxz7QT/Ycb/5AVQTi/vMnls87/p4P2KTV1NR6nYWiqrlOAR8LRItIeadpApDVwA8X4yOvrk3nkk7tuKZfrK+jGGMaoC5nocSKSLRzPxy4DtiDp5BPd1abCbzXVCFN49t99Cwb03O4c2xPgmzINGMCUl1OO4gHFjr94EHAW6r6gYjsBlJF5NfAVuClJsxpGtlr69IJCw1iRpINmWZMoKq1gKvqdmBENfMPAqOaIpRpWnmFpSzZmsV/jehOVLtQX8cxxjSQfROzFXp78xGKS93cMSbR11GMMV6wAt7KuN3Kq2szGJXYiUHdOvg6jjHGC1bAW5lVX53k8JlC7rzCvrhjTKCzAt7KLFybTpfItnx7cFdfRzHGeMkKeCuSfuocK788yW2jexIabL96YwKd/RW3Iq+tyyA0WPjeaDt10JiWwAp4K3G+THlr0xEmD4mnS6QNmWZMS2AFvJVYm11GfnEZM+3gpTEthhXwVkBV+TijlMHdOjDyEhsyzZiWwgp4K7Dh0BkyC5SZY23INGNaEivgLVx+cSm//GA37UNhyjAbMs2YlsQKeAtWXOri3lc38eWxfH4wtC3hbYJ9HckY04hsEMQWqszl5sdvbGX9oTM8e8twonL3+TqSMaaRWQu8BVJV5r6zg+W7j/PklMFMHd7d15GMMU3ACngLo6r8z9I9LN6cyUPX9mXmFYm+jmSMaSJ1GZGnh4ikichuEdklIg868zuJyHIR2ef8tPPT/MC8VQd4cc0hZl2RyIMT+/o6jjGmCdWlBV4GPKqqg4AxwBwRGQTMBVaoal9ghTNtfOiNDYf5/YdfMnV4N/77hkF2yqAxLVytBVxVs1V1i3M/H894mN2BqcBCZ7WFwLSmCmlqt3RHNo8v2UFK/1iemTHMxrk0phWoVx+4iCTiGV5tPRCnqtnOomNAXKMmM3W2Zt9JHkzdyshLOvL8bUl2pUFjWglR1bqtKBIBrAJ+o6rviEiuqkZXWp6jqhf0g4vIbGA2QFxcXFJqamqDghYUFBAREdGgbZuDr/IdzHXxu43FdGkXxNxRYbQPrb7lbfvPO5bPO5bPOykpKZtVNfmCBapa6w0IBf4DPFJp3pdAvHM/HviytsdJSkrShkpLS2vwts3BF/m+OnZWhz31Hx33u0/0eF7RRde1/ecdy+cdy+cdYJNWU1PrchaKAC8Be1T1T5UWvQ/MdO7PBN5r6KeLqb/MnELueGkDocFB/PPu0XTpYJeINaa1qcs3Ma8E7gB2iMg2Z97PgaeBt0TkbiADuLlpIpqqThWc586XNlBYUsabPxjLJZ3b+TqSMcYHai3gqvopUNMpDRMbN46pTX5xKbMWbOBoXhH/vHs0A+NtZHljWis7XSGAFJe6uGfhJvZm5zPv9iSSEzv5OpIxxofsYlYBovziVBvSPRenSunfxdeRjDE+Zi3wAOB2f31xqqdutItTGWM8rID7uXPny/jpv7azeHMmD1/bjzvHJvo6kjHGT1gXih/bcjiHR97cRsaZQh6Y2JcHJvbxdSRjjB+xAu6HSl1u/u8n+3kubT9dO4SReu8YRvfu7OtYxhg/YwXczxw8WcDDb27ji8w8bhrZnSdvHEyHsFBfxzLG+CEr4H5CVVm0/jC/+X97aBsaxPO3jeT6y+J9HcsY48esgPuBE/nFzP3XDj7Ze4JxfWN4ZsYw4uyr8caYWlgB97GPdh1j7js7OHe+jCenDOLOsYl2LW9jTJ1YAfeRgvNl/Orfu3lz0xEGd+vAs7cMp29cpK9jGWMCiBVwH9ickcPDb27jSE4hPxx/KQ9d2482IXZKvjGmfqyAN6NSl5u/rtjHc2n76RYdzls/GMvldj0TY0wDWQFvJgec0wO3Z+YxPSmBJ6YMItJODzTGeMEKeBNxu5X00+fYnpnHtiO5pG48THhoMPNuG8lkOz3QGNMIrIA3AlXlVJGbpTuy2Z6Zx/bMXHZk5ZFfXAZAWGgQ4/t14ZdTB9vIOcaYRlNrAReRl4EbgBOqOsSZ1wl4E0gE0oGbVTWn6WL6lxP5xWw/ksf2LKdYZ+Zx+lwJsIXQYGFA1w7cOKwbQxOiGJoQTd8uEYTYSPHGmEZWlxb4K8DfgFcrzZsLrFDVp0VkrjP9WOPHq59Sl5vs3GIycwvJzCkiM6eIrJwiMnMKOVdSRpvgINqEBNEmJJg2wUG0DXGmK+ZXma40/8y5Er44ksv2zDyOnS0GIEigb5dIJgzoQljRCaanXM6A+EjahgT7eE8YY1qDugyptlpEEqvMngqMd+4vBFbSDAX8fJmLo7nFFUU5M6eIrNyv7x8/W4xbv15fBLp2CKN7dDhdIsMoKXNTUuYmr6jUue+ixOWumF9S5qbE5abUpdU+f6+Y9ozq1YmhCVEM6xHNoPgOtG/r2YUrV65kWI/opt4FxhhTQTwj1teykqeAf1CpCyVXVaOd+wLklE9Xs+1sYDZAXFxcUmpqar1DLtx1ns3HSzlb8s1vKAYJdGwrxIQLMeFBxIQLnSvd7xQmhDTgW41uVcrcVNxK3UpYiNA+tObHKigoICIiot7P1Vwsn3csn3csn3dSUlI2q2py1fleH8RUVRWRGj8FVHU+MB8gOTlZx48fX+/n2KX7KXMfIHlgbxI6htO9YzgJHcPp2iHMb/qWV65cSUNeW3OxfN6xfN6xfE2joQX8uIjEq2q2iMQDJxozVFVzUvowWDIZP75vUz6NMcYElIY2X98HZjr3ZwLvNU4cY4wxdVVrAReRN4C1QH8RyRSRu4GngetEZB9wrTNtjDGmGdXlLJTv1bBoYiNnMcYYUw/+cQTQGGNMvVkBN8aYAGUF3BhjApQVcGOMCVBWwI0xJkDV6av0jfZkIieBjAZuHgOcasQ4jc3yecfyecfyecff8/VU1diqM5u1gHtDRDZVdy0Af2H5vGP5vGP5vOPv+WpiXSjGGBOgrIAbY0yACqQCPt/XAWph+bxj+bxj+bzj7/mqFTB94MYYY74pkFrgxhhjKrECbowxAcrvC7iI/EFE9orIdhFZIiLRlZb9TET2i8iXIvJtH+WbISK7RMQtIsmV5ieKSJGIbHNuL/hTPmeZz/dflTxPikhWpX12va8zAYjIJGcf7XcG8fYrIpIuIjucfbbJD/K8LCInRGRnpXmdRGS5iOxzfnb0s3x++d6rjd8XcGA5MERVhwJfAT8DEJFBwK3AYGAS8LyI+GI4+J3ATcDqapYdUNXhzu2+Zs5Vrtp8frT/qvpzpX221NdhnH3yHDAZGAR8z9l3/ibF2Wf+cC7zK3jeU5XNBVaoal9ghTPtK69wYT7ws/deXfh9AVfVj1S1zJlcByQ496cCqap6XlUPAfuBUT7It0dVv2zu562ri+Tzi/0XAEYB+1X1oKqWAKl49p2pgaquBs5UmT0VWOjcXwhMa9ZQldSQLyD5fQGv4n8By5z73YEjlZZlOvP8SS8R2Soiq0RknK/DVOGv++9HTnfZy778N7sSf91PlSnwkYhsFpHZvg5TgzhVzXbuHwPifBmmBv723quV16PSNwYR+RjoWs2ix1X1PWedx4EyYFFzZnOeu9Z81cgGLlHV0yKSBLwrIoNV9ayf5POJi2UF5gG/wlOQfgX8Ec+Htrm4q1Q1S0S6AMtFZK/TyvRLqqoi4m/nLwfke88vCriqXnux5SIyC7gBmKhfn7ieBfSotFqCM6/Z89WwzXngvHN/s4gcAPoBjX6QqSH5aMb9V1lds4rIi8AHTRynLnyyn+pDVbOcnydEZAmebh9/K+DHRSReVbNFJB444etAlanq8fL7fvTeq5Xfd6GIyCTgp8CNqlpYadH7wK0i0lZEegF9gQ2+yFgdEYktPygoIr3x5Dvo21Tf4Hf7z/nDLvdfeA7A+tpGoK+I9BKRNngO/L7v40wVRKS9iESW3we+hX/st6reB2Y692cC/vafoT++92qnqn59w3Nw7Qiwzbm9UGnZ48AB4Etgso/y/ReeftHzwHHgP8787wK7nMxbgCn+lM9f9l+VrK8BO4DteP7g432dycl1PZ4zoA7g6ZbyeaZK2XoDXzi3Xf6QD3gDTxdiqfPeuxvojOfsk33Ax0AnP8vnl++92m72VXpjjAlQft+FYowxpnpWwI0xJkBZATfGmABlBdwYYwKUFXBjjAlQVsBNiyUijztXYtzuXGFutIisrHzFPhFJFpGVzv3xIpLnrLtXRJ7xWXhj6sAKuGmRRGQsnm/vjlTPlSyv5etrmnQRkck1bLpGVYcDI4AbROTKpk9rTMNYATctVTxwSj2XNEBVT6nqUWfZH/B8ialGqlqE50tY3QFE5JpK14reWv7tR2N8yQq4aak+AnqIyFci8ryIXFNp2VqgRERSatrYuRpdX76+pshPgDlO63wcUNREuY2pMyvgpkVS1QIgCZgNnATedC6KVu7XwP+pZtNxIvIFngtW/UdVjznzPwP+JCIPANH69TXqjfEZK+CmxVJVl6quVNUngB/huT5N+bJPgHBgTJXN1qjqMDwjFd0tIsOd9Z8G7nG2+UxEBjTHazDmYqyAmxZJRPqLSN9Ks4YDGVVW+zWeK11eQD2jFD0NPOY83qWqukNVf4fnCoVWwI3PWQE3LVUEsFBEdovIdjzjWT5ZeQX1jHt48iKP8QJwtYgkAg+JyE7nsUr5emQoY3zGrkZojDEBylrgxhgToKyAG2NMgLICbowxAcoKuDHGBCgr4MYYE6CsgBtjTICyAm6MMQHq/wPSnJPr9c5I7gAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pcjOBppfjV_1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "feat_file='/content/gdrive/My Drive/EE5611_project_features/base_scrnn_features.h5'\n",
        "fid=h5py.File(feat_file,'r');\n",
        "x_tr = fid['x_tr']\n",
        "y_tr = fid['y_tr']\n",
        "x_ts=fid['x_ts']\n",
        "y_ts=fid['y_ts']\n",
        "snr_ts = np.array(fid['snr_ts'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z3UcqcfkP1MQ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "3c85daf1-4eb0-45cb-90be-ff56fea34d88"
      },
      "source": [
        "model_dir = '/content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/'\n",
        "m_baseline_scrnn_tanh = baseline_scrnn_tanh()\n",
        "checkpoint = keras.callbacks.ModelCheckpoint(model_dir+ \"model_epoch-{epoch:02d}_val_acc-{val_accuracy:.2f}.h5\", monitor='val_accuracy', verbose=2, save_best_only=False, save_weights_only=False, mode='auto', period=1)\n",
        "history = m_baseline_scrnn_tanh.fit(x=x_tr,\n",
        "       y=y_tr,\n",
        "       validation_data=[x_ts,y_ts],\n",
        "       batch_size=400,\n",
        "       epochs=50,\n",
        "       shuffle='batch',\n",
        "       callbacks = [checkpoint],\n",
        "       verbose=1)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 128, 2)            0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_2 (Batch (None, 128, 2)            8         \n",
            "_________________________________________________________________\n",
            "conv1d_3 (Conv1D)            (None, 124, 128)          1408      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_2 (MaxPooling1 (None, 41, 128)           0         \n",
            "_________________________________________________________________\n",
            "conv1d_4 (Conv1D)            (None, 37, 128)           82048     \n",
            "_________________________________________________________________\n",
            "lstm_3 (LSTM)                (None, 37, 128)           131584    \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 37, 128)           0         \n",
            "_________________________________________________________________\n",
            "lstm_4 (LSTM)                (None, 37, 128)           131584    \n",
            "_________________________________________________________________\n",
            "dropout_4 (Dropout)          (None, 37, 128)           0         \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 4736)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 11)                52107     \n",
            "=================================================================\n",
            "Total params: 398,739\n",
            "Trainable params: 398,735\n",
            "Non-trainable params: 4\n",
            "_________________________________________________________________\n",
            "Train on 198000 samples, validate on 22000 samples\n",
            "Epoch 1/50\n",
            "198000/198000 [==============================] - 34s 174us/step - loss: 1.6102 - accuracy: 0.3765 - val_loss: 2.1100 - val_accuracy: 0.2615\n",
            "\n",
            "Epoch 00001: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-01_val_acc-0.26.h5\n",
            "Epoch 2/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 1.3164 - accuracy: 0.4888 - val_loss: 1.2569 - val_accuracy: 0.5188\n",
            "\n",
            "Epoch 00002: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-02_val_acc-0.52.h5\n",
            "Epoch 3/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 1.2349 - accuracy: 0.5235 - val_loss: 1.1885 - val_accuracy: 0.5455\n",
            "\n",
            "Epoch 00003: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-03_val_acc-0.55.h5\n",
            "Epoch 4/50\n",
            "198000/198000 [==============================] - 25s 127us/step - loss: 1.1819 - accuracy: 0.5455 - val_loss: 1.1559 - val_accuracy: 0.5570\n",
            "\n",
            "Epoch 00004: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-04_val_acc-0.56.h5\n",
            "Epoch 5/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 1.1561 - accuracy: 0.5557 - val_loss: 1.1421 - val_accuracy: 0.5652\n",
            "\n",
            "Epoch 00005: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-05_val_acc-0.57.h5\n",
            "Epoch 6/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 1.1396 - accuracy: 0.5614 - val_loss: 1.1301 - val_accuracy: 0.5646\n",
            "\n",
            "Epoch 00006: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-06_val_acc-0.56.h5\n",
            "Epoch 7/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 1.1275 - accuracy: 0.5665 - val_loss: 1.1204 - val_accuracy: 0.5640\n",
            "\n",
            "Epoch 00007: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-07_val_acc-0.56.h5\n",
            "Epoch 8/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.1180 - accuracy: 0.5716 - val_loss: 1.1192 - val_accuracy: 0.5725\n",
            "\n",
            "Epoch 00008: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-08_val_acc-0.57.h5\n",
            "Epoch 9/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 1.1122 - accuracy: 0.5751 - val_loss: 1.1042 - val_accuracy: 0.5831\n",
            "\n",
            "Epoch 00009: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-09_val_acc-0.58.h5\n",
            "Epoch 10/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.1047 - accuracy: 0.5805 - val_loss: 1.1131 - val_accuracy: 0.5803\n",
            "\n",
            "Epoch 00010: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-10_val_acc-0.58.h5\n",
            "Epoch 11/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 1.0955 - accuracy: 0.5877 - val_loss: 1.0948 - val_accuracy: 0.5911\n",
            "\n",
            "Epoch 00011: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-11_val_acc-0.59.h5\n",
            "Epoch 12/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 1.0837 - accuracy: 0.5956 - val_loss: 1.0856 - val_accuracy: 0.5985\n",
            "\n",
            "Epoch 00012: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-12_val_acc-0.60.h5\n",
            "Epoch 13/50\n",
            "198000/198000 [==============================] - 23s 116us/step - loss: 1.0736 - accuracy: 0.6027 - val_loss: 1.0829 - val_accuracy: 0.6009\n",
            "\n",
            "Epoch 00013: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-13_val_acc-0.60.h5\n",
            "Epoch 14/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 1.0630 - accuracy: 0.6086 - val_loss: 1.0682 - val_accuracy: 0.6083\n",
            "\n",
            "Epoch 00014: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-14_val_acc-0.61.h5\n",
            "Epoch 15/50\n",
            "198000/198000 [==============================] - 24s 124us/step - loss: 1.0558 - accuracy: 0.6127 - val_loss: 1.0710 - val_accuracy: 0.6101\n",
            "\n",
            "Epoch 00015: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-15_val_acc-0.61.h5\n",
            "Epoch 16/50\n",
            "198000/198000 [==============================] - 25s 124us/step - loss: 1.0483 - accuracy: 0.6170 - val_loss: 1.0666 - val_accuracy: 0.6134\n",
            "\n",
            "Epoch 00016: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-16_val_acc-0.61.h5\n",
            "Epoch 17/50\n",
            "198000/198000 [==============================] - 25s 128us/step - loss: 1.0395 - accuracy: 0.6214 - val_loss: 1.0620 - val_accuracy: 0.6132\n",
            "\n",
            "Epoch 00017: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-17_val_acc-0.61.h5\n",
            "Epoch 18/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 1.0330 - accuracy: 0.6241 - val_loss: 1.0665 - val_accuracy: 0.6127\n",
            "\n",
            "Epoch 00018: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-18_val_acc-0.61.h5\n",
            "Epoch 19/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 1.0274 - accuracy: 0.6265 - val_loss: 1.0585 - val_accuracy: 0.6136\n",
            "\n",
            "Epoch 00019: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-19_val_acc-0.61.h5\n",
            "Epoch 20/50\n",
            "198000/198000 [==============================] - 24s 124us/step - loss: 1.0183 - accuracy: 0.6309 - val_loss: 1.0665 - val_accuracy: 0.6120\n",
            "\n",
            "Epoch 00020: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-20_val_acc-0.61.h5\n",
            "Epoch 21/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.0118 - accuracy: 0.6336 - val_loss: 1.0657 - val_accuracy: 0.6149\n",
            "\n",
            "Epoch 00021: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-21_val_acc-0.61.h5\n",
            "Epoch 22/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 1.0051 - accuracy: 0.6367 - val_loss: 1.0800 - val_accuracy: 0.6112\n",
            "\n",
            "Epoch 00022: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-22_val_acc-0.61.h5\n",
            "Epoch 23/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9986 - accuracy: 0.6401 - val_loss: 1.0726 - val_accuracy: 0.6173\n",
            "\n",
            "Epoch 00023: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-23_val_acc-0.62.h5\n",
            "Epoch 24/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9916 - accuracy: 0.6432 - val_loss: 1.0666 - val_accuracy: 0.6159\n",
            "\n",
            "Epoch 00024: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-24_val_acc-0.62.h5\n",
            "Epoch 25/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 0.9825 - accuracy: 0.6465 - val_loss: 1.0810 - val_accuracy: 0.6099\n",
            "\n",
            "Epoch 00025: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-25_val_acc-0.61.h5\n",
            "Epoch 26/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.9750 - accuracy: 0.6490 - val_loss: 1.0747 - val_accuracy: 0.6180\n",
            "\n",
            "Epoch 00026: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-26_val_acc-0.62.h5\n",
            "Epoch 27/50\n",
            "198000/198000 [==============================] - 25s 125us/step - loss: 0.9687 - accuracy: 0.6518 - val_loss: 1.0893 - val_accuracy: 0.6148\n",
            "\n",
            "Epoch 00027: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-27_val_acc-0.61.h5\n",
            "Epoch 28/50\n",
            "198000/198000 [==============================] - 24s 123us/step - loss: 0.9603 - accuracy: 0.6557 - val_loss: 1.0897 - val_accuracy: 0.6155\n",
            "\n",
            "Epoch 00028: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-28_val_acc-0.62.h5\n",
            "Epoch 29/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.9532 - accuracy: 0.6580 - val_loss: 1.1036 - val_accuracy: 0.6170\n",
            "\n",
            "Epoch 00029: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-29_val_acc-0.62.h5\n",
            "Epoch 30/50\n",
            "198000/198000 [==============================] - 24s 122us/step - loss: 0.9428 - accuracy: 0.6626 - val_loss: 1.1092 - val_accuracy: 0.6116\n",
            "\n",
            "Epoch 00030: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-30_val_acc-0.61.h5\n",
            "Epoch 31/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.9368 - accuracy: 0.6650 - val_loss: 1.1130 - val_accuracy: 0.6146\n",
            "\n",
            "Epoch 00031: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-31_val_acc-0.61.h5\n",
            "Epoch 32/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.9285 - accuracy: 0.6678 - val_loss: 1.1076 - val_accuracy: 0.6123\n",
            "\n",
            "Epoch 00032: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-32_val_acc-0.61.h5\n",
            "Epoch 33/50\n",
            "198000/198000 [==============================] - 24s 120us/step - loss: 0.9205 - accuracy: 0.6711 - val_loss: 1.1307 - val_accuracy: 0.6133\n",
            "\n",
            "Epoch 00033: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-33_val_acc-0.61.h5\n",
            "Epoch 34/50\n",
            "198000/198000 [==============================] - 23s 119us/step - loss: 0.9138 - accuracy: 0.6742 - val_loss: 1.1387 - val_accuracy: 0.6102\n",
            "\n",
            "Epoch 00034: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-34_val_acc-0.61.h5\n",
            "Epoch 35/50\n",
            "198000/198000 [==============================] - 24s 119us/step - loss: 0.9032 - accuracy: 0.6778 - val_loss: 1.1487 - val_accuracy: 0.6120\n",
            "\n",
            "Epoch 00035: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-35_val_acc-0.61.h5\n",
            "Epoch 36/50\n",
            "198000/198000 [==============================] - 23s 115us/step - loss: 0.8933 - accuracy: 0.6813 - val_loss: 1.1539 - val_accuracy: 0.6110\n",
            "\n",
            "Epoch 00036: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-36_val_acc-0.61.h5\n",
            "Epoch 37/50\n",
            "198000/198000 [==============================] - 23s 117us/step - loss: 0.8888 - accuracy: 0.6829 - val_loss: 1.1608 - val_accuracy: 0.6087\n",
            "\n",
            "Epoch 00037: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-37_val_acc-0.61.h5\n",
            "Epoch 38/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.8788 - accuracy: 0.6862 - val_loss: 1.1625 - val_accuracy: 0.6103\n",
            "\n",
            "Epoch 00038: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-38_val_acc-0.61.h5\n",
            "Epoch 39/50\n",
            "198000/198000 [==============================] - 23s 117us/step - loss: 0.8702 - accuracy: 0.6904 - val_loss: 1.1726 - val_accuracy: 0.6104\n",
            "\n",
            "Epoch 00039: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-39_val_acc-0.61.h5\n",
            "Epoch 40/50\n",
            "198000/198000 [==============================] - 24s 120us/step - loss: 0.8629 - accuracy: 0.6927 - val_loss: 1.1890 - val_accuracy: 0.6104\n",
            "\n",
            "Epoch 00040: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-40_val_acc-0.61.h5\n",
            "Epoch 41/50\n",
            "198000/198000 [==============================] - 23s 118us/step - loss: 0.8540 - accuracy: 0.6963 - val_loss: 1.1954 - val_accuracy: 0.6039\n",
            "\n",
            "Epoch 00041: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-41_val_acc-0.60.h5\n",
            "Epoch 42/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.8460 - accuracy: 0.6986 - val_loss: 1.2054 - val_accuracy: 0.6081\n",
            "\n",
            "Epoch 00042: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-42_val_acc-0.61.h5\n",
            "Epoch 43/50\n",
            "198000/198000 [==============================] - 23s 118us/step - loss: 0.8377 - accuracy: 0.7007 - val_loss: 1.2229 - val_accuracy: 0.6088\n",
            "\n",
            "Epoch 00043: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-43_val_acc-0.61.h5\n",
            "Epoch 44/50\n",
            "198000/198000 [==============================] - 23s 119us/step - loss: 0.8308 - accuracy: 0.7036 - val_loss: 1.2102 - val_accuracy: 0.6063\n",
            "\n",
            "Epoch 00044: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-44_val_acc-0.61.h5\n",
            "Epoch 45/50\n",
            "198000/198000 [==============================] - 24s 121us/step - loss: 0.8248 - accuracy: 0.7058 - val_loss: 1.2351 - val_accuracy: 0.5997\n",
            "\n",
            "Epoch 00045: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-45_val_acc-0.60.h5\n",
            "Epoch 46/50\n",
            "198000/198000 [==============================] - 25s 126us/step - loss: 0.8158 - accuracy: 0.7097 - val_loss: 1.2447 - val_accuracy: 0.6017\n",
            "\n",
            "Epoch 00046: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-46_val_acc-0.60.h5\n",
            "Epoch 47/50\n",
            "198000/198000 [==============================] - 23s 118us/step - loss: 0.8077 - accuracy: 0.7120 - val_loss: 1.2394 - val_accuracy: 0.6042\n",
            "\n",
            "Epoch 00047: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-47_val_acc-0.60.h5\n",
            "Epoch 48/50\n",
            "198000/198000 [==============================] - 23s 118us/step - loss: 0.7996 - accuracy: 0.7149 - val_loss: 1.2567 - val_accuracy: 0.5996\n",
            "\n",
            "Epoch 00048: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-48_val_acc-0.60.h5\n",
            "Epoch 49/50\n",
            "198000/198000 [==============================] - 23s 117us/step - loss: 0.7902 - accuracy: 0.7178 - val_loss: 1.2626 - val_accuracy: 0.6044\n",
            "\n",
            "Epoch 00049: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-49_val_acc-0.60.h5\n",
            "Epoch 50/50\n",
            "198000/198000 [==============================] - 23s 119us/step - loss: 0.7841 - accuracy: 0.7204 - val_loss: 1.2712 - val_accuracy: 0.6003\n",
            "\n",
            "Epoch 00050: saving model to /content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/model_epoch-50_val_acc-0.60.h5\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1J3KzqkbQAOx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 769
        },
        "outputId": "b0b4c704-5109-40e3-98ca-744965bde9e0"
      },
      "source": [
        "acc_baseline_scrnn_tanh = history.history['accuracy']\n",
        "val_acc_baseline_scrnn_tanh = history.history['val_accuracy']\n",
        "loss_baseline_scrnn_tanh = history.history['loss']\n",
        "val_loss_baseline_scrnn_tanh = history.history['val_loss']\n",
        "epochs = range(1, len(acc_baseline_scrnn_tanh) + 1)\n",
        "\n",
        "print('===================================================')\n",
        "print('Baseline scrnn tanh\\n\\n')\n",
        "print('acc_baseline_scrnn_tanh\\n',acc_baseline_scrnn_tanh)\n",
        "print('val_acc_baseline_scrnn_tanh\\n',val_acc_baseline_scrnn_tanh)\n",
        "print('loss_baseline_scrnn_tanh\\n',loss_baseline_scrnn_tanh)\n",
        "print('val_loss_baseline_scrnn_tanh\\n',val_loss_baseline_scrnn_tanh)\n",
        "\n",
        "plt.plot(epochs, acc_baseline_scrnn_tanh, 'bo', label='Training acc')\n",
        "plt.plot(epochs, val_acc_baseline_scrnn_tanh, 'b', label='Validation acc')\n",
        "plt.title('Training and validation accuracy for Baseline SCRNN tanh')\n",
        "plt.legend()\n",
        "plt.figure()\n",
        "plt.plot(epochs, loss_baseline_scrnn_tanh, 'bo', label='Training loss')\n",
        "plt.plot(epochs, val_loss_baseline_scrnn_tanh, 'b', label='Validation loss')\n",
        "plt.title('Training and validation loss for Baseline Scrnn tanh')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "===================================================\n",
            "Baseline scrnn tanh\n",
            "\n",
            "\n",
            "acc_baseline_scrnn_tanh\n",
            " [0.37645453, 0.48882827, 0.5234646, 0.54550505, 0.5556667, 0.56144947, 0.5664899, 0.57158583, 0.5750808, 0.58046967, 0.58774745, 0.5955606, 0.6026667, 0.60855556, 0.61267173, 0.6170101, 0.62144446, 0.62408084, 0.6264596, 0.6308788, 0.6336465, 0.6366717, 0.6400909, 0.6431515, 0.6464949, 0.6490151, 0.6517677, 0.65566164, 0.6579747, 0.6625505, 0.6650303, 0.6678333, 0.6710656, 0.67415655, 0.67784345, 0.6812828, 0.6829091, 0.6862121, 0.690399, 0.6927121, 0.6962778, 0.69856566, 0.7007475, 0.70361114, 0.70584345, 0.7096818, 0.71195453, 0.71488386, 0.7178182, 0.7204394]\n",
            "val_acc_baseline_scrnn_tanh\n",
            " [0.2614545524120331, 0.5187727212905884, 0.5454545617103577, 0.5569999814033508, 0.5652272701263428, 0.5645909309387207, 0.5639545321464539, 0.5724999904632568, 0.5830909013748169, 0.5802727341651917, 0.5910909175872803, 0.5985000133514404, 0.6008636355400085, 0.6082727313041687, 0.6100909113883972, 0.6134091019630432, 0.6132272481918335, 0.6127272844314575, 0.6136363744735718, 0.6120454668998718, 0.6149091124534607, 0.61122727394104, 0.6173181533813477, 0.6158636212348938, 0.6098636388778687, 0.6179545521736145, 0.6147727370262146, 0.6154545545578003, 0.6169999837875366, 0.6116363406181335, 0.6146363615989685, 0.6122727394104004, 0.6133182048797607, 0.6101818084716797, 0.6120454668998718, 0.6110454797744751, 0.6087272763252258, 0.6102727055549622, 0.6104090809822083, 0.6103636622428894, 0.6038636565208435, 0.6081363558769226, 0.6087727546691895, 0.6063181757926941, 0.5997272729873657, 0.6016818284988403, 0.6041818261146545, 0.5995908975601196, 0.6043636202812195, 0.6002727150917053]\n",
            "loss_baseline_scrnn_tanh\n",
            " [1.6101957311533919, 1.316390791083827, 1.2348934915330676, 1.1819468401899242, 1.1560888735934942, 1.1395572840565382, 1.1274934476072138, 1.117971947096815, 1.1122481310006345, 1.1046545187632242, 1.095495345736995, 1.0837316426363859, 1.0735808523014339, 1.0629714210828145, 1.055752828988162, 1.0483094416483485, 1.0395126967719106, 1.0329707528605605, 1.0273690277879888, 1.01825005297709, 1.0117856922775808, 1.0051111033468536, 0.9986231312607274, 0.9915955453207999, 0.982475077383446, 0.9749569104175375, 0.9687184161610074, 0.9603281714699485, 0.9532016880584485, 0.9427674930505078, 0.936765974459022, 0.9285034454230107, 0.9204915300764219, 0.9137541944330388, 0.9032061244502212, 0.8932858966817759, 0.8888433665940256, 0.8788392022402599, 0.8701650443703237, 0.86286004745599, 0.8539692288697368, 0.8459971340015681, 0.8376509985538444, 0.8307770724248404, 0.8247697690520623, 0.8158102284778248, 0.8076977501011858, 0.7996108306778802, 0.7902187106585262, 0.784055485027005]\n",
            "val_loss_baseline_scrnn_tanh\n",
            " [2.11003187678077, 1.2569285200400786, 1.1884517989375374, 1.1559289184398949, 1.1420565206652762, 1.1300627502358773, 1.12040040822361, 1.1192366807361727, 1.104210699910552, 1.1130900006594444, 1.0948006980020595, 1.0856399935305077, 1.0828892660142049, 1.0682410517474636, 1.071004867823682, 1.0666159631442331, 1.0620296024528877, 1.0664861500190785, 1.058539195366426, 1.0665159934560176, 1.065709530215893, 1.0800092310351002, 1.072605390195994, 1.0666080372247961, 1.0809789703502743, 1.0746840120236316, 1.0893304979896956, 1.0897097378101699, 1.1035976962881582, 1.1092040355204187, 1.1130208736138931, 1.107637202814228, 1.1307145398245086, 1.138745266777352, 1.1486755288665749, 1.1539366489215932, 1.160832222918163, 1.1625065313381293, 1.1726190750653174, 1.1889763632141694, 1.19535111803912, 1.2054275180286245, 1.2228938217642116, 1.2102109612973766, 1.2351146981445873, 1.2446722043328091, 1.239392276582095, 1.2567033934470435, 1.2625965299591453, 1.2712006510973715]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "31z57S1Mgc_m",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "607ac8a0-dd64-419a-f4fa-458e4407d19e"
      },
      "source": [
        "# baseline scrnn tanh model evaluation\n",
        "baseline_scrnn_tanh_model = utils.evaluate_model('/content/gdrive/My Drive/EE5611_project_features/SCRNN_models/Baseline_SCRNN_tanh_models/','model_epoch-26_val_acc-0.62.h5',np.array(x_ts),np.array(y_ts),snr_ts)\n",
        "print(baseline_scrnn_tanh_model)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "===============================================================\n",
            "\n",
            "SNR: -20dB, Validation Accuracy: 9.636363636363637\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.06      0.01      0.02       100\n",
            "           1       0.00      0.00      0.00       100\n",
            "           2       0.17      0.02      0.04       100\n",
            "           3       0.12      0.02      0.03       100\n",
            "           4       0.20      0.01      0.02       100\n",
            "           5       0.33      0.01      0.02       100\n",
            "           6       0.10      0.96      0.17       100\n",
            "           7       0.07      0.01      0.02       100\n",
            "           8       0.11      0.01      0.02       100\n",
            "           9       0.25      0.01      0.02       100\n",
            "          10       0.00      0.00      0.00       100\n",
            "\n",
            "    accuracy                           0.10      1100\n",
            "   macro avg       0.13      0.10      0.03      1100\n",
            "weighted avg       0.13      0.10      0.03      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 1  2  1  1  0  0 90  5  0  0  0]\n",
            " [ 3  0  0  0  0  0 96  0  0  0  1]\n",
            " [ 1  0  2  1  1  0 91  1  1  0  2]\n",
            " [ 2  4  2  2  1  0 86  1  2  0  0]\n",
            " [ 1  1  1  2  1  0 92  0  0  1  1]\n",
            " [ 2  0  1  3  0  1 88  1  3  0  1]\n",
            " [ 2  0  0  0  0  1 96  1  0  0  0]\n",
            " [ 2  0  0  3  0  0 92  1  0  1  1]\n",
            " [ 2  1  2  1  1  1 88  2  1  0  1]\n",
            " [ 1  2  0  1  0  0 92  2  1  1  0]\n",
            " [ 0  1  3  3  1  0 90  0  1  1  0]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -18dB, Validation Accuracy: 10.0\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.00      0.00      0.00       100\n",
            "           1       0.00      0.00      0.00       100\n",
            "           2       0.17      0.03      0.05       100\n",
            "           3       0.24      0.04      0.07       100\n",
            "           4       0.33      0.01      0.02       100\n",
            "           5       0.33      0.01      0.02       100\n",
            "           6       0.09      0.93      0.17       100\n",
            "           7       0.17      0.04      0.06       100\n",
            "           8       0.20      0.01      0.02       100\n",
            "           9       0.10      0.01      0.02       100\n",
            "          10       0.17      0.02      0.04       100\n",
            "\n",
            "    accuracy                           0.10      1100\n",
            "   macro avg       0.16      0.10      0.04      1100\n",
            "weighted avg       0.16      0.10      0.04      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 0  2  1  0  0  0 93  2  0  1  1]\n",
            " [ 2  0  0  1  0  0 94  1  0  0  2]\n",
            " [ 1  2  3  1  0  0 89  0  1  2  1]\n",
            " [ 1  1  2  4  0  0 90  2  0  0  0]\n",
            " [ 0  0  2  0  1  0 92  1  2  1  1]\n",
            " [ 1  4  6  1  1  1 74  5  0  3  4]\n",
            " [ 0  2  0  1  0  0 93  4  0  0  0]\n",
            " [ 2  0  0  3  1  0 87  4  0  2  1]\n",
            " [ 2  0  3  1  0  1 91  1  1  0  0]\n",
            " [ 4  0  0  4  0  0 88  3  0  1  0]\n",
            " [ 2  2  1  1  0  1 89  1  1  0  2]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -16dB, Validation Accuracy: 9.454545454545455\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.00      0.00      0.00       100\n",
            "           1       0.00      0.00      0.00       100\n",
            "           2       0.10      0.02      0.03       100\n",
            "           3       0.15      0.03      0.05       100\n",
            "           4       0.08      0.01      0.02       100\n",
            "           5       0.40      0.02      0.04       100\n",
            "           6       0.10      0.90      0.17       100\n",
            "           7       0.00      0.00      0.00       100\n",
            "           8       0.14      0.03      0.05       100\n",
            "           9       0.25      0.02      0.04       100\n",
            "          10       0.08      0.01      0.02       100\n",
            "\n",
            "    accuracy                           0.09      1100\n",
            "   macro avg       0.12      0.09      0.04      1100\n",
            "weighted avg       0.12      0.09      0.04      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 0  1  3  0  1  0 91  4  0  0  0]\n",
            " [ 1  0  0  1  0  0 93  2  1  2  0]\n",
            " [ 1  2  2  1  2  0 89  0  2  1  0]\n",
            " [ 2  0  4  3  0  1 81  3  2  3  1]\n",
            " [ 0  0  1  2  1  0 89  2  3  0  2]\n",
            " [ 7 10  8  2  2  2 45 10  8  0  6]\n",
            " [ 1  1  0  1  2  0 90  2  1  0  2]\n",
            " [ 1  1  1  2  0  0 95  0  0  0  0]\n",
            " [ 1  2  0  2  1  1 87  2  3  0  1]\n",
            " [ 0  0  2  5  0  1 88  1  1  2  0]\n",
            " [ 0  0  0  1  3  0 94  0  1  0  1]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -14dB, Validation Accuracy: 11.727272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.08      0.03      0.04       100\n",
            "           1       0.00      0.00      0.00       100\n",
            "           2       0.11      0.03      0.05       100\n",
            "           3       0.28      0.10      0.15       100\n",
            "           4       0.10      0.01      0.02       100\n",
            "           5       0.93      0.14      0.24       100\n",
            "           6       0.11      0.92      0.19       100\n",
            "           7       0.00      0.00      0.00       100\n",
            "           8       0.10      0.03      0.05       100\n",
            "           9       0.29      0.02      0.04       100\n",
            "          10       0.04      0.01      0.02       100\n",
            "\n",
            "    accuracy                           0.12      1100\n",
            "   macro avg       0.19      0.12      0.07      1100\n",
            "weighted avg       0.19      0.12      0.07      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 3  5  1  0  1  0 88  1  1  0  0]\n",
            " [ 2  0  1  0  3  0 94  0  0  0  0]\n",
            " [ 5  1  3  2  1  0 83  0  2  1  2]\n",
            " [ 0  0  1 10  0  0 80  0  3  3  3]\n",
            " [ 3  0  2  3  1  0 89  0  1  0  1]\n",
            " [10 16 11  2  1 14  2 18 18  0  8]\n",
            " [ 3  2  0  2  0  0 92  0  0  0  1]\n",
            " [ 0  0  2  2  1  0 92  0  1  0  2]\n",
            " [ 7  5  5  6  0  1 63  6  3  0  4]\n",
            " [ 2  2  1  8  1  0 81  2  0  2  1]\n",
            " [ 1  0  1  1  1  0 92  2  0  1  1]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -12dB, Validation Accuracy: 16.181818181818183\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.04      0.01      0.02       100\n",
            "           1       0.04      0.02      0.03       100\n",
            "           2       0.40      0.19      0.26       100\n",
            "           3       0.46      0.28      0.35       100\n",
            "           4       0.15      0.02      0.04       100\n",
            "           5       0.76      0.22      0.34       100\n",
            "           6       0.12      0.88      0.21       100\n",
            "           7       0.05      0.03      0.04       100\n",
            "           8       0.19      0.10      0.13       100\n",
            "           9       0.20      0.03      0.05       100\n",
            "          10       0.00      0.00      0.00       100\n",
            "\n",
            "    accuracy                           0.16      1100\n",
            "   macro avg       0.22      0.16      0.13      1100\n",
            "weighted avg       0.22      0.16      0.13      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 1  2  1  3  0  0 85  5  1  0  2]\n",
            " [ 0  2  0  0  1  1 91  1  3  0  1]\n",
            " [ 5  3 19  5  2  3 52  3  4  2  2]\n",
            " [ 0  0  3 28  1  0 64  1  0  3  0]\n",
            " [ 1  2  4  2  2  0 84  1  0  2  2]\n",
            " [ 0 20  1  0  0 22  0 24 30  0  3]\n",
            " [ 0  1  3  1  2  0 88  1  3  1  0]\n",
            " [ 0  0  2  2  2  0 87  3  0  3  1]\n",
            " [11 12  9  1  2  3 28 18 10  1  5]\n",
            " [ 3  3  4 14  1  0 67  1  1  3  3]\n",
            " [ 2  2  2  5  0  0 89  0  0  0  0]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -10dB, Validation Accuracy: 23.727272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.09      0.04      0.06       100\n",
            "           1       0.06      0.02      0.03       100\n",
            "           2       0.61      0.51      0.55       100\n",
            "           3       0.44      0.40      0.42       100\n",
            "           4       0.33      0.08      0.13       100\n",
            "           5       0.65      0.36      0.46       100\n",
            "           6       0.16      0.95      0.28       100\n",
            "           7       0.06      0.04      0.05       100\n",
            "           8       0.20      0.15      0.17       100\n",
            "           9       0.17      0.04      0.07       100\n",
            "          10       0.11      0.02      0.03       100\n",
            "\n",
            "    accuracy                           0.24      1100\n",
            "   macro avg       0.26      0.24      0.20      1100\n",
            "weighted avg       0.26      0.24      0.20      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 4  6  8  1  2  1 76  2  0  0  0]\n",
            " [ 3  2  2  0  2  1 78  5  4  1  2]\n",
            " [21  3 51  0  1  0 13  3  5  2  1]\n",
            " [ 2  0  1 40  3  0 39  2  2 10  1]\n",
            " [ 2  1  2  6  8  0 73  1  3  3  1]\n",
            " [ 0  5  1  0  0 36  1 17 40  0  0]\n",
            " [ 0  0  1  2  0  0 95  1  0  1  0]\n",
            " [ 2  0  4  1  1  0 84  4  2  1  1]\n",
            " [ 3 12  9  2  3 16  6 25 15  0  9]\n",
            " [ 2  0  3 37  2  0 47  1  2  4  2]\n",
            " [ 4  3  2  2  2  1 78  3  2  1  2]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -8dB, Validation Accuracy: 36.09090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.22      0.13      0.16       100\n",
            "           1       0.16      0.07      0.10       100\n",
            "           2       0.56      0.74      0.64       100\n",
            "           3       0.41      0.61      0.49       100\n",
            "           4       0.45      0.28      0.35       100\n",
            "           5       0.52      0.49      0.50       100\n",
            "           6       0.30      0.92      0.45       100\n",
            "           7       0.16      0.10      0.12       100\n",
            "           8       0.27      0.29      0.28       100\n",
            "           9       0.38      0.19      0.25       100\n",
            "          10       0.44      0.15      0.22       100\n",
            "\n",
            "    accuracy                           0.36      1100\n",
            "   macro avg       0.35      0.36      0.32      1100\n",
            "weighted avg       0.35      0.36      0.32      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[13  3 21  1  0  0 50  6  4  0  2]\n",
            " [10  7 12  2  2  4 35 14  9  2  3]\n",
            " [17  3 74  1  1  0  1  1  1  1  0]\n",
            " [ 0  0  0 61  9  1  7  1  0 20  1]\n",
            " [ 3  2  3 15 28  4 30  2  3  5  5]\n",
            " [ 0  2  0  0  0 49  1  4 44  0  0]\n",
            " [ 2  0  1  2  0  0 92  1  2  0  0]\n",
            " [ 7  4  9  1  5  0 50 10 10  1  3]\n",
            " [ 0 14  2  0  0 31  1 20 29  0  3]\n",
            " [ 1  0  1 61 10  1  4  0  1 19  2]\n",
            " [ 6  8  8  4  7  5 36  5  4  2 15]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -6dB, Validation Accuracy: 55.81818181818182\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.62      0.41      0.49       100\n",
            "           1       0.43      0.22      0.29       100\n",
            "           2       0.65      0.94      0.77       100\n",
            "           3       0.52      0.85      0.64       100\n",
            "           4       0.70      0.73      0.71       100\n",
            "           5       0.55      0.70      0.61       100\n",
            "           6       0.71      0.93      0.81       100\n",
            "           7       0.32      0.23      0.27       100\n",
            "           8       0.42      0.52      0.46       100\n",
            "           9       0.48      0.22      0.30       100\n",
            "          10       0.58      0.39      0.47       100\n",
            "\n",
            "    accuracy                           0.56      1100\n",
            "   macro avg       0.54      0.56      0.53      1100\n",
            "weighted avg       0.54      0.56      0.53      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[41  1 35  1  2  1 11  4  2  1  1]\n",
            " [ 8 22  5  1  1 12  7 21 14  0  9]\n",
            " [ 4  0 94  1  0  1  0  0  0  0  0]\n",
            " [ 0  0  0 85  2  0  0  0  0 13  0]\n",
            " [ 1  1  2  4 73  1  2  1  1  9  5]\n",
            " [ 0  0  0  0  0 70  0  0 30  0  0]\n",
            " [ 0  1  2  3  0  0 93  0  1  0  0]\n",
            " [ 6 16  5  2  1  4 12 23 17  1 13]\n",
            " [ 0  2  0  1  0 35  2  8 52  0  0]\n",
            " [ 1  0  1 63 13  0  0  0  0 22  0]\n",
            " [ 5  8  1  3 13  4  4 16  7  0 39]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -4dB, Validation Accuracy: 68.9090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.92      0.65      0.76       100\n",
            "           1       0.60      0.46      0.52       100\n",
            "           2       0.74      0.97      0.84       100\n",
            "           3       0.54      0.85      0.66       100\n",
            "           4       0.87      0.89      0.88       100\n",
            "           5       0.58      0.83      0.69       100\n",
            "           6       0.94      0.96      0.95       100\n",
            "           7       0.45      0.39      0.42       100\n",
            "           8       0.64      0.66      0.65       100\n",
            "           9       0.51      0.24      0.33       100\n",
            "          10       0.83      0.68      0.75       100\n",
            "\n",
            "    accuracy                           0.69      1100\n",
            "   macro avg       0.69      0.69      0.68      1100\n",
            "weighted avg       0.69      0.69      0.68      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[65  3 30  0  0  0  0  1  1  0  0]\n",
            " [ 2 46  0  0  0  9  2 27  7  2  5]\n",
            " [ 1  0 97  0  0  0  2  0  0  0  0]\n",
            " [ 0  0  0 85  1  0  0  0  0 14  0]\n",
            " [ 0  0  0  2 89  0  0  1  0  7  1]\n",
            " [ 0  0  0  0  0 83  0  1 16  0  0]\n",
            " [ 0  1  2  0  0  0 96  1  0  0  0]\n",
            " [ 3 22  0  0  2 14  2 39 10  0  8]\n",
            " [ 0  0  1  0  0 29  0  4 66  0  0]\n",
            " [ 0  0  0 68  8  0  0  0  0 24  0]\n",
            " [ 0  5  1  1  2  7  0 13  3  0 68]]\n",
            "===============================================================\n",
            "\n",
            "SNR: -2dB, Validation Accuracy: 80.27272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.90      0.94       100\n",
            "           1       0.74      0.68      0.71       100\n",
            "           2       0.91      0.97      0.94       100\n",
            "           3       0.54      0.92      0.68       100\n",
            "           4       0.98      0.94      0.96       100\n",
            "           5       0.73      0.84      0.78       100\n",
            "           6       0.96      0.97      0.97       100\n",
            "           7       0.71      0.59      0.64       100\n",
            "           8       0.74      0.83      0.78       100\n",
            "           9       0.70      0.23      0.35       100\n",
            "          10       0.96      0.96      0.96       100\n",
            "\n",
            "    accuracy                           0.80      1100\n",
            "   macro avg       0.81      0.80      0.79      1100\n",
            "weighted avg       0.81      0.80      0.79      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[90  0  9  1  0  0  0  0  0  0  0]\n",
            " [ 1 68  0  0  0  9  1 18  3  0  0]\n",
            " [ 0  0 97  0  0  0  2  1  0  0  0]\n",
            " [ 0  0  0 92  0  0  0  0  0  8  0]\n",
            " [ 0  1  0  2 94  0  0  0  0  2  1]\n",
            " [ 0  0  0  0  0 84  0  2 14  0  0]\n",
            " [ 1  1  1  0  0  0 97  0  0  0  0]\n",
            " [ 0 20  0  0  0  6  1 59 12  0  2]\n",
            " [ 0  0  0  0  0 15  0  2 83  0  0]\n",
            " [ 0  0  0 74  2  0  0  0  0 23  1]\n",
            " [ 0  2  0  0  0  1  0  1  0  0 96]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 0dB, Validation Accuracy: 87.72727272727273\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.98      0.98       100\n",
            "           1       0.91      0.98      0.94       100\n",
            "           2       0.97      0.98      0.98       100\n",
            "           3       0.58      0.97      0.73       100\n",
            "           4       0.96      0.98      0.97       100\n",
            "           5       0.84      0.80      0.82       100\n",
            "           6       0.99      0.94      0.96       100\n",
            "           7       0.96      0.86      0.91       100\n",
            "           8       0.78      0.86      0.82       100\n",
            "           9       0.89      0.31      0.46       100\n",
            "          10       1.00      0.99      0.99       100\n",
            "\n",
            "    accuracy                           0.88      1100\n",
            "   macro avg       0.90      0.88      0.87      1100\n",
            "weighted avg       0.90      0.88      0.87      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[98  0  2  0  0  0  0  0  0  0  0]\n",
            " [ 0 98  0  0  0  0  0  2  0  0  0]\n",
            " [ 1  0 98  0  0  0  1  0  0  0  0]\n",
            " [ 0  0  0 97  0  0  0  0  0  3  0]\n",
            " [ 0  0  0  1 98  0  0  0  0  1  0]\n",
            " [ 0  0  0  0  0 80  0  0 20  0  0]\n",
            " [ 0  0  1  2  0  1 94  1  1  0  0]\n",
            " [ 0 10  0  0  0  1  0 86  3  0  0]\n",
            " [ 0  0  0  0  0 13  0  1 86  0  0]\n",
            " [ 0  0  0 66  3  0  0  0  0 31  0]\n",
            " [ 0  0  0  0  1  0  0  0  0  0 99]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 2dB, Validation Accuracy: 89.45454545454545\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.97      0.98       100\n",
            "           1       0.97      0.99      0.98       100\n",
            "           2       0.98      0.98      0.98       100\n",
            "           3       0.59      0.95      0.73       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.90      0.86      0.88       100\n",
            "           6       0.92      0.92      0.92       100\n",
            "           7       0.99      0.95      0.97       100\n",
            "           8       0.86      0.88      0.87       100\n",
            "           9       0.85      0.34      0.49       100\n",
            "          10       0.97      1.00      0.99       100\n",
            "\n",
            "    accuracy                           0.89      1100\n",
            "   macro avg       0.91      0.89      0.89      1100\n",
            "weighted avg       0.91      0.89      0.89      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 97   0   0   0   0   0   3   0   0   0   0]\n",
            " [  0  99   0   0   0   0   1   0   0   0   0]\n",
            " [  0   0  98   1   0   0   1   0   0   0   0]\n",
            " [  0   0   0  95   0   0   0   0   0   5   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  86   0   0  14   0   0]\n",
            " [  0   1   2   1   0   0  92   0   0   1   3]\n",
            " [  1   2   0   0   0   1   1  95   0   0   0]\n",
            " [  0   0   0   0   0   9   2   1  88   0   0]\n",
            " [  0   0   0  65   1   0   0   0   0  34   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 4dB, Validation Accuracy: 90.63636363636364\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.96      0.99      0.98       100\n",
            "           1       0.99      0.99      0.99       100\n",
            "           2       1.00      0.99      0.99       100\n",
            "           3       0.59      0.99      0.74       100\n",
            "           4       1.00      0.99      0.99       100\n",
            "           5       0.93      0.88      0.90       100\n",
            "           6       0.92      0.92      0.92       100\n",
            "           7       1.00      0.96      0.98       100\n",
            "           8       0.88      0.91      0.89       100\n",
            "           9       0.95      0.35      0.51       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.91      1100\n",
            "   macro avg       0.93      0.91      0.90      1100\n",
            "weighted avg       0.93      0.91      0.90      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   0   0   0   0   1   0   0   0   0]\n",
            " [  0  99   0   0   0   0   1   0   0   0   0]\n",
            " [  0   0  99   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0  99   0   0   0   0   0   1   0]\n",
            " [  0   0   0   0  99   0   0   0   0   1   0]\n",
            " [  0   0   0   0   0  88   0   0  12   0   0]\n",
            " [  4   0   0   3   0   1  92   0   0   0   0]\n",
            " [  0   1   0   0   0   0   2  96   1   0   0]\n",
            " [  0   0   0   0   0   6   3   0  91   0   0]\n",
            " [  0   0   0  65   0   0   0   0   0  35   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 6dB, Validation Accuracy: 91.81818181818181\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.97      1.00      0.99       100\n",
            "           1       0.99      0.97      0.98       100\n",
            "           2       0.96      1.00      0.98       100\n",
            "           3       0.64      1.00      0.78       100\n",
            "           4       1.00      1.00      1.00       100\n",
            "           5       0.92      0.87      0.89       100\n",
            "           6       0.96      0.92      0.94       100\n",
            "           7       0.99      0.97      0.98       100\n",
            "           8       0.89      0.94      0.91       100\n",
            "           9       1.00      0.43      0.60       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.92      1100\n",
            "   macro avg       0.94      0.92      0.91      1100\n",
            "weighted avg       0.94      0.92      0.91      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0  97   0   0   0   2   1   0   0   0   0]\n",
            " [  0   0 100   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  1   0   0   0   0  87   2   0  10   0   0]\n",
            " [  2   1   4   0   0   0  92   1   0   0   0]\n",
            " [  0   0   0   0   0   0   1  97   2   0   0]\n",
            " [  0   0   0   0   0   6   0   0  94   0   0]\n",
            " [  0   0   0  57   0   0   0   0   0  43   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 8dB, Validation Accuracy: 92.0909090909091\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00       100\n",
            "           1       1.00      0.99      0.99       100\n",
            "           2       0.97      0.99      0.98       100\n",
            "           3       0.65      1.00      0.78       100\n",
            "           4       1.00      1.00      1.00       100\n",
            "           5       0.91      0.90      0.90       100\n",
            "           6       0.95      0.90      0.92       100\n",
            "           7       0.96      0.99      0.98       100\n",
            "           8       0.91      0.89      0.90       100\n",
            "           9       0.98      0.47      0.64       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.92      1100\n",
            "   macro avg       0.94      0.92      0.92      1100\n",
            "weighted avg       0.94      0.92      0.92      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0  99   0   0   0   0   1   0   0   0   0]\n",
            " [  0   0  99   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   1   0   0  90   0   0   9   0   0]\n",
            " [  1   0   2   2   0   0  90   4   0   1   0]\n",
            " [  0   0   0   0   0   0   1  99   0   0   0]\n",
            " [  0   0   0   0   0   9   2   0  89   0   0]\n",
            " [  0   0   0  53   0   0   0   0   0  47   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 10dB, Validation Accuracy: 92.9090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.99      0.99       100\n",
            "           1       1.00      0.99      0.99       100\n",
            "           2       1.00      0.99      0.99       100\n",
            "           3       0.65      1.00      0.79       100\n",
            "           4       1.00      1.00      1.00       100\n",
            "           5       0.92      0.93      0.93       100\n",
            "           6       0.95      0.96      0.96       100\n",
            "           7       0.98      0.99      0.99       100\n",
            "           8       0.91      0.90      0.90       100\n",
            "           9       1.00      0.47      0.64       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.93      1100\n",
            "   macro avg       0.95      0.93      0.93      1100\n",
            "weighted avg       0.95      0.93      0.93      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   0   0   0   0   1   0   0   0   0]\n",
            " [  0  99   0   0   0   0   1   0   0   0   0]\n",
            " [  0   0  99   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  93   0   0   7   0   0]\n",
            " [  0   0   0   1   0   0  96   1   2   0   0]\n",
            " [  0   0   0   0   0   0   1  99   0   0   0]\n",
            " [  0   0   0   0   0   8   1   1  90   0   0]\n",
            " [  0   0   0  53   0   0   0   0   0  47   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 12dB, Validation Accuracy: 92.18181818181819\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.98      0.98      0.98       100\n",
            "           1       1.00      1.00      1.00       100\n",
            "           2       0.95      0.98      0.97       100\n",
            "           3       0.65      1.00      0.79       100\n",
            "           4       1.00      1.00      1.00       100\n",
            "           5       0.95      0.86      0.90       100\n",
            "           6       0.96      0.92      0.94       100\n",
            "           7       0.98      0.99      0.99       100\n",
            "           8       0.88      0.93      0.90       100\n",
            "           9       0.96      0.48      0.64       100\n",
            "          10       1.00      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.92      1100\n",
            "   macro avg       0.94      0.92      0.92      1100\n",
            "weighted avg       0.94      0.92      0.92      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 98   0   1   0   0   0   1   0   0   0   0]\n",
            " [  0 100   0   0   0   0   0   0   0   0   0]\n",
            " [  1   0  98   0   0   0   1   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  86   0   1  13   0   0]\n",
            " [  1   0   4   1   0   0  92   0   0   2   0]\n",
            " [  0   0   0   0   0   0   1  99   0   0   0]\n",
            " [  0   0   0   0   0   5   1   1  93   0   0]\n",
            " [  0   0   0  52   0   0   0   0   0  48   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 14dB, Validation Accuracy: 91.9090909090909\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.98      0.99       100\n",
            "           1       0.98      0.98      0.98       100\n",
            "           2       0.99      1.00      1.00       100\n",
            "           3       0.65      1.00      0.78       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.94      0.84      0.89       100\n",
            "           6       0.93      0.94      0.94       100\n",
            "           7       0.99      0.98      0.98       100\n",
            "           8       0.85      0.93      0.89       100\n",
            "           9       1.00      0.46      0.63       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.92      1100\n",
            "   macro avg       0.94      0.92      0.92      1100\n",
            "weighted avg       0.94      0.92      0.92      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 98   0   0   0   0   0   2   0   0   0   0]\n",
            " [  0  98   0   0   0   0   2   0   0   0   0]\n",
            " [  0   0 100   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  84   0   0  16   0   0]\n",
            " [  0   1   1   1   1   0  94   1   0   0   1]\n",
            " [  0   1   0   0   0   0   1  98   0   0   0]\n",
            " [  0   0   0   0   0   5   2   0  93   0   0]\n",
            " [  0   0   0  54   0   0   0   0   0  46   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 16dB, Validation Accuracy: 93.27272727272727\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      0.99      0.99       100\n",
            "           1       1.00      1.00      1.00       100\n",
            "           2       0.98      1.00      0.99       100\n",
            "           3       0.65      1.00      0.79       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.96      0.91      0.93       100\n",
            "           6       0.96      0.96      0.96       100\n",
            "           7       0.99      1.00      1.00       100\n",
            "           8       0.93      0.93      0.93       100\n",
            "           9       1.00      0.47      0.64       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.93      1100\n",
            "   macro avg       0.95      0.93      0.93      1100\n",
            "weighted avg       0.95      0.93      0.93      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[ 99   0   0   0   0   0   1   0   0   0   0]\n",
            " [  0 100   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0 100   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  91   2   0   7   0   0]\n",
            " [  1   0   1   0   1   0  96   0   0   0   1]\n",
            " [  0   0   0   0   0   0   0 100   0   0   0]\n",
            " [  0   0   1   0   0   4   1   1  93   0   0]\n",
            " [  0   0   0  53   0   0   0   0   0  47   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "===============================================================\n",
            "\n",
            "SNR: 18dB, Validation Accuracy: 92.0909090909091\n",
            "\n",
            "\n",
            "             Classification Report\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.99      1.00      1.00       100\n",
            "           1       1.00      1.00      1.00       100\n",
            "           2       0.99      1.00      1.00       100\n",
            "           3       0.60      1.00      0.75       100\n",
            "           4       0.99      1.00      1.00       100\n",
            "           5       0.98      0.88      0.93       100\n",
            "           6       0.98      0.93      0.95       100\n",
            "           7       1.00      1.00      1.00       100\n",
            "           8       0.89      0.97      0.93       100\n",
            "           9       0.97      0.35      0.51       100\n",
            "          10       0.99      1.00      1.00       100\n",
            "\n",
            "    accuracy                           0.92      1100\n",
            "   macro avg       0.94      0.92      0.91      1100\n",
            "weighted avg       0.94      0.92      0.91      1100\n",
            "\n",
            "\n",
            "             Confusion Matrix\n",
            "[[100   0   0   0   0   0   0   0   0   0   0]\n",
            " [  0 100   0   0   0   0   0   0   0   0   0]\n",
            " [  0   0 100   0   0   0   0   0   0   0   0]\n",
            " [  0   0   0 100   0   0   0   0   0   0   0]\n",
            " [  0   0   0   0 100   0   0   0   0   0   0]\n",
            " [  0   0   0   0   0  88   1   0  11   0   0]\n",
            " [  1   0   1   1   1   0  93   0   1   1   1]\n",
            " [  0   0   0   0   0   0   0 100   0   0   0]\n",
            " [  0   0   0   0   0   2   1   0  97   0   0]\n",
            " [  0   0   0  65   0   0   0   0   0  35   0]\n",
            " [  0   0   0   0   0   0   0   0   0   0 100]]\n",
            "[9.636363636363637, 10.0, 9.454545454545455, 11.727272727272727, 16.181818181818183, 23.727272727272727, 36.09090909090909, 55.81818181818182, 68.9090909090909, 80.27272727272727, 87.72727272727273, 89.45454545454545, 90.63636363636364, 91.81818181818181, 92.0909090909091, 92.9090909090909, 92.18181818181819, 91.9090909090909, 93.27272727272727, 92.0909090909091]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zuiqdzccWOZl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}