Malaria detection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Problem Statement\n",
    "Malaria is a life-threatening disease caused by parasites that are transmitted to people through the bites of infected female Anopheles mosquitoes. It is preventable and curable.\n",
    "\n",
    "Malaria must be recognized promptly in order to treat the patient in time and to prevent further spread of infection in the community via local mosquitoes.\n",
    "Malaria should be considered a potential medical emergency and should be treated accordingly. Delay in diagnosis and treatment is a leading cause of death in malaria patients in the United States.\n",
    "\n",
    "Diagnosis of malaria can be difficult:\n",
    "\n",
    "* here malaria is not endemic any more (such as in the United States), health-care providers may not be familiar with the disease. Clinicians seeing a malaria patient may forget to consider malaria among the potential diagnoses and not order the needed diagnostic tests. Laboratorians may lack experience with malaria and fail to detect parasites when examining blood smears under the microscope.\n",
    "* In some malaria-endemic areas, malaria transmission is so intense that a large proportion of the population is infected but not made ill by the parasites. Such carriers have developed just enough immunity to protect them from malarial illness but not from malarial infection. In that situation, finding malaria parasites in an ill person does not necessarily mean that the illness is caused by the parasites.\n",
    "\n",
    "(source: [cdc.gov](cdc.gov/malaria/diagnosis_treatment/diagnosis.html))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With given cell images dataset we try to classife whether patient is ill or not. \n",
    "Dataset may be devided into two separate sets: parasitized cells and uninfected cells. Its a simple classification problem with 2 labels. \n",
    "Images with parasite will be marked as 1 and uninfected as 0. We use CNN to perform a classification. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import cv2\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "from PIL import Image\n",
    "from keras import backend, optimizers\n",
    "from keras.layers import Activation, BatchNormalization, Conv2D, Dense, Dropout, GlobalAveragePooling2D, Flatten, MaxPooling2D\n",
    "from keras.models import Sequential\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.utils import np_utils\n",
    "from sklearn.model_selection import train_test_split, KFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "parasitized_dir = 'data/cell_images/Parasitized'\n",
    "uninfected_dir = 'data/cell_images/Uninfected'\n",
    "parasitized_cells = os.listdir(parasitized_dir)\n",
    "uninfected_cells = os.listdir(uninfected_dir)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_examples(photo_names: list, photo_dir: str) -> None:\n",
    "    plt.figure(figsize = (15,15))\n",
    "    for num, img_name in enumerate(photo_names):\n",
    "        plt.subplot(1, len(photo_names), num + 1)\n",
    "        img_path = os.path.join(photo_dir, img_name)\n",
    "        pil_im = Image.open(img_path, 'r')\n",
    "        plt.imshow(pil_im)\n",
    "        plt.tight_layout()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Parasitized cells example\n",
    "Visualization of of infected and uninfected cells images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "example_parasitized_cells = parasitized_cells[:6]\n",
    "\n",
    "plot_examples(example_parasitized_cells, parasitized_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Uninfected cells example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 6 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "example_uninfected_cells = uninfected_cells[:6]\n",
    "\n",
    "plot_examples(example_uninfected_cells, uninfected_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see quite significant differences at the pictures. Parasitized cell have visible swings on the surface. These differences are good work for CNNs."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Data preprocessing\n",
    "Data preprocessing and labeling is pretty straightforward. Everything what I have done is resizing to 50x50 images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess_images(images: list, base_path: str, img_label: int, resize: tuple = (50, 50)) -> tuple:\n",
    "    X, y = [], []\n",
    "    for image in images:\n",
    "        try:\n",
    "            img_arr = plt.imread(os.path.join(base_path, image))\n",
    "            resized_img = cv2.resize(img_arr, resize)\n",
    "            y.append(img_label)\n",
    "            X.append(resized_img)\n",
    "        except OSError as exc:\n",
    "            print(exc)\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cannot identify image file 'data/cell_images/Parasitized\\\\Thumbs.db'\n",
      "cannot identify image file 'data/cell_images/Uninfected\\\\Thumbs.db'\n"
     ]
    }
   ],
   "source": [
    "X_p, y_p = preprocess_images(parasitized_cells, parasitized_dir, 1)\n",
    "X_u, y_u = preprocess_images(uninfected_cells, uninfected_dir, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.array([*X_p, *X_u])\n",
    "y = np.array([*y_p, *y_u])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Splitting data in proporion 7:3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_train = np_utils.to_categorical(y_train, num_classes = 2)\n",
    "y_test = np_utils.to_categorical(y_test, num_classes = 2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "CNN built for image recognition is a simple CNN with two CONV => RELU => POOL => CONV => RELU layers and single fully connected layer. I've also used BatchNormalization layer after each RELU activation making our network train faster, converage more quickly and making activation function more viable. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNNNet:\n",
    "    @staticmethod\n",
    "    def build(width: int, height: int, depth: int, classes: int):\n",
    "        model = Sequential()\n",
    "        input_shape = (height, width, depth)\n",
    "        channel_dimension = -1\n",
    "        \n",
    "        if backend.image_data_format() == \"channels_first\":\n",
    "            input_shape = (depth, height, width)\n",
    "            channel_dimension = 1\n",
    "            \n",
    "        # first CONV => RELU => POOL => CONV => RELU layer set\n",
    "        model.add(Conv2D(32, (3, 3), input_shape=input_shape))\n",
    "        model.add(Activation(\"relu\"))\n",
    "        model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "        model.add(BatchNormalization(axis=channel_dimension))\n",
    "        model.add(Conv2D(32, (3, 3), input_shape=input_shape))\n",
    "        model.add(Activation(\"relu\"))\n",
    "        model.add(BatchNormalization(axis=channel_dimension))\n",
    "        model.add(Dropout(0.2))\n",
    "\n",
    "        # second CONV => RELU => POOL => CONV => RELU layer set\n",
    "        model.add(Conv2D(32, (3, 3), input_shape=input_shape))\n",
    "        model.add(Activation(\"relu\"))\n",
    "        model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "        model.add(BatchNormalization(axis=channel_dimension))\n",
    "        model.add(Conv2D(32, (3, 3), input_shape=input_shape))\n",
    "        model.add(Activation(\"relu\"))\n",
    "        model.add(BatchNormalization(axis=channel_dimension))\n",
    "        model.add(Dropout(0.2))\n",
    "\n",
    "        # first (and only) set of FC => RELU layers\n",
    "        model.add(Flatten())\n",
    "        model.add(Dense(64))\n",
    "        model.add(Activation(\"relu\"))\n",
    "        model.add(BatchNormalization(axis=channel_dimension))\n",
    "        model.add(Dropout(0.5))\n",
    "        \n",
    "        # softmax classifier\n",
    "        model.add(Dense(classes))\n",
    "        model.add(Activation(\"softmax\"))\n",
    "        \n",
    "        return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Data augmentation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "CNN needs a lot of data. Thanks tu Data Augmentaion process we can genarete enough data for our model. I've tried simple data augmentation (small zooms, small rotations and pretty small horizontal/vertical shifts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_image_generator = ImageDataGenerator(\n",
    "#         rescale=1/255.,\n",
    "        fill_mode='nearest',  # fill points outside the boundaries \n",
    "        featurewise_center=False,  # set input mean to 0 over the dataset\n",
    "        samplewise_center=False,  # set each sample mean to 0\n",
    "        featurewise_std_normalization=False,  # divide inputs by std of the dataset\n",
    "        samplewise_std_normalization=False,  # divide each input by its std\n",
    "        zca_whitening=False,  # apply ZCA whitening\n",
    "        rotation_range=30,  # randomly rotate images in the range (degrees, 0 to 180)\n",
    "        zoom_range = 0.3, # Randomly zoom image \n",
    "        width_shift_range=0.2,  # randomly shift images horizontally (fraction of total width)\n",
    "        height_shift_range=0.2,  # randomly shift images vertically (fraction of total height)\n",
    "        horizontal_flip=True,  # randomly flip images\n",
    "        vertical_flip=False  # randomly flip images\n",
    ")  \n",
    "\n",
    "validation_image_generator = ImageDataGenerator(\n",
    "#     rescale=1/255.\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data = train_image_generator.flow(X_train, y_train, batch_size=64, shuffle=True)\n",
    "validation_data = validation_image_generator.flow(X_test, y_test, batch_size=64, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 48, 48, 32)        896       \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 48, 48, 32)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_1 (MaxPooling2 (None, 24, 24, 32)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_1 (Batch (None, 24, 24, 32)        128       \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 22, 22, 32)        9248      \n",
      "_________________________________________________________________\n",
      "activation_2 (Activation)    (None, 22, 22, 32)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_2 (Batch (None, 22, 22, 32)        128       \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 22, 22, 32)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 20, 20, 32)        9248      \n",
      "_________________________________________________________________\n",
      "activation_3 (Activation)    (None, 20, 20, 32)        0         \n",
      "_________________________________________________________________\n",
      "max_pooling2d_2 (MaxPooling2 (None, 10, 10, 32)        0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_3 (Batch (None, 10, 10, 32)        128       \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 8, 8, 32)          9248      \n",
      "_________________________________________________________________\n",
      "activation_4 (Activation)    (None, 8, 8, 32)          0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_4 (Batch (None, 8, 8, 32)          128       \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 8, 8, 32)          0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 2048)              0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 64)                131136    \n",
      "_________________________________________________________________\n",
      "activation_5 (Activation)    (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_5 (Batch (None, 64)                256       \n",
      "_________________________________________________________________\n",
      "dropout_3 (Dropout)          (None, 64)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 2)                 130       \n",
      "_________________________________________________________________\n",
      "activation_6 (Activation)    (None, 2)                 0         \n",
      "=================================================================\n",
      "Total params: 160,674\n",
      "Trainable params: 160,290\n",
      "Non-trainable params: 384\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "height, width, depth = 50, 50, 3\n",
    "classes = 2\n",
    "\n",
    "model = CNNNet.build(width, height, depth, classes)\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "steps_per_epoch = X_train.shape[0] // 64\n",
    "adam_optimizer = optimizers.Adam(lr = 0.001, decay = 0.001 / 64)\n",
    "model.compile(loss='categorical_crossentropy', optimizer=adam_optimizer, metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1454 - accuracy: 0.9489 - val_loss: 0.0367 - val_accuracy: 0.9609\n",
      "Epoch 2/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1495 - accuracy: 0.9459 - val_loss: 0.0333 - val_accuracy: 0.9583\n",
      "Epoch 3/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1509 - accuracy: 0.9437 - val_loss: 0.0500 - val_accuracy: 0.9505\n",
      "Epoch 4/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1450 - accuracy: 0.9487 - val_loss: 0.1277 - val_accuracy: 0.9479\n",
      "Epoch 5/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1525 - accuracy: 0.9465 - val_loss: 0.1571 - val_accuracy: 0.9557\n",
      "Epoch 6/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1445 - accuracy: 0.9475 - val_loss: 0.1140 - val_accuracy: 0.9583\n",
      "Epoch 7/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1513 - accuracy: 0.9476 - val_loss: 0.0728 - val_accuracy: 0.9583\n",
      "Epoch 8/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1482 - accuracy: 0.9500 - val_loss: 0.1145 - val_accuracy: 0.9766\n",
      "Epoch 9/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1476 - accuracy: 0.9472 - val_loss: 0.0511 - val_accuracy: 0.9714\n",
      "Epoch 10/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1494 - accuracy: 0.9464 - val_loss: 0.1542 - val_accuracy: 0.9688\n",
      "Epoch 11/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1449 - accuracy: 0.9474 - val_loss: 0.1785 - val_accuracy: 0.9557\n",
      "Epoch 12/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1509 - accuracy: 0.9456 - val_loss: 0.1704 - val_accuracy: 0.9740\n",
      "Epoch 13/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1519 - accuracy: 0.9450 - val_loss: 0.1062 - val_accuracy: 0.9766\n",
      "Epoch 14/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1498 - accuracy: 0.9459 - val_loss: 0.0541 - val_accuracy: 0.9583\n",
      "Epoch 15/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1523 - accuracy: 0.9464 - val_loss: 0.1838 - val_accuracy: 0.9583\n",
      "Epoch 16/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1459 - accuracy: 0.9482 - val_loss: 0.0460 - val_accuracy: 0.9688\n",
      "Epoch 17/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1529 - accuracy: 0.9461 - val_loss: 0.0491 - val_accuracy: 0.9688\n",
      "Epoch 18/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1406 - accuracy: 0.9484 - val_loss: 0.1687 - val_accuracy: 0.9583\n",
      "Epoch 19/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1505 - accuracy: 0.9463 - val_loss: 0.1675 - val_accuracy: 0.9531\n",
      "Epoch 20/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1450 - accuracy: 0.9478 - val_loss: 0.1476 - val_accuracy: 0.9661\n",
      "Epoch 21/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1483 - accuracy: 0.9484 - val_loss: 0.1476 - val_accuracy: 0.9740\n",
      "Epoch 22/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1492 - accuracy: 0.9456 - val_loss: 0.0635 - val_accuracy: 0.9428\n",
      "Epoch 23/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1456 - accuracy: 0.9484 - val_loss: 0.0150 - val_accuracy: 0.9740\n",
      "Epoch 24/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1447 - accuracy: 0.9482 - val_loss: 0.1536 - val_accuracy: 0.9479\n",
      "Epoch 25/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1489 - accuracy: 0.9459 - val_loss: 0.0909 - val_accuracy: 0.9583\n",
      "Epoch 26/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1472 - accuracy: 0.9450 - val_loss: 0.0786 - val_accuracy: 0.9609\n",
      "Epoch 27/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1452 - accuracy: 0.9490 - val_loss: 0.1753 - val_accuracy: 0.9531\n",
      "Epoch 28/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1499 - accuracy: 0.9459 - val_loss: 0.1571 - val_accuracy: 0.9505\n",
      "Epoch 29/50\n",
      "301/301 [==============================] - 11s 38ms/step - loss: 0.1416 - accuracy: 0.9492 - val_loss: 0.0179 - val_accuracy: 0.9688\n",
      "Epoch 30/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1460 - accuracy: 0.9480 - val_loss: 0.1008 - val_accuracy: 0.9609\n",
      "Epoch 31/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1480 - accuracy: 0.9467 - val_loss: 0.0827 - val_accuracy: 0.9740\n",
      "Epoch 32/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1518 - accuracy: 0.9467 - val_loss: 0.1625 - val_accuracy: 0.9635\n",
      "Epoch 33/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1473 - accuracy: 0.9474 - val_loss: 0.0584 - val_accuracy: 0.9766\n",
      "Epoch 34/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1483 - accuracy: 0.9465 - val_loss: 0.1513 - val_accuracy: 0.9740\n",
      "Epoch 35/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1480 - accuracy: 0.9480 - val_loss: 0.0702 - val_accuracy: 0.9740\n",
      "Epoch 36/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1455 - accuracy: 0.9488 - val_loss: 0.1319 - val_accuracy: 0.9714\n",
      "Epoch 37/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1379 - accuracy: 0.9504 - val_loss: 0.0217 - val_accuracy: 0.9635\n",
      "Epoch 38/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1461 - accuracy: 0.9465 - val_loss: 0.0836 - val_accuracy: 0.9740\n",
      "Epoch 39/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1482 - accuracy: 0.9467 - val_loss: 0.1820 - val_accuracy: 0.9401\n",
      "Epoch 40/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1413 - accuracy: 0.9493 - val_loss: 0.1719 - val_accuracy: 0.9609\n",
      "Epoch 41/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1418 - accuracy: 0.9497 - val_loss: 0.0999 - val_accuracy: 0.9714\n",
      "Epoch 42/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1426 - accuracy: 0.9491 - val_loss: 0.1089 - val_accuracy: 0.9661\n",
      "Epoch 43/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1451 - accuracy: 0.9476 - val_loss: 0.3609 - val_accuracy: 0.9557\n",
      "Epoch 44/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1490 - accuracy: 0.9464 - val_loss: 0.0944 - val_accuracy: 0.9578\n",
      "Epoch 45/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1438 - accuracy: 0.9488 - val_loss: 0.1190 - val_accuracy: 0.9531\n",
      "Epoch 46/50\n",
      "301/301 [==============================] - 11s 37ms/step - loss: 0.1366 - accuracy: 0.9517 - val_loss: 0.0542 - val_accuracy: 0.9557\n",
      "Epoch 47/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1435 - accuracy: 0.9479 - val_loss: 0.1192 - val_accuracy: 0.9792\n",
      "Epoch 48/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1428 - accuracy: 0.9486 - val_loss: 0.0644 - val_accuracy: 0.9427\n",
      "Epoch 49/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1541 - accuracy: 0.9463 - val_loss: 0.0404 - val_accuracy: 0.9740\n",
      "Epoch 50/50\n",
      "301/301 [==============================] - 11s 36ms/step - loss: 0.1380 - accuracy: 0.9487 - val_loss: 0.2690 - val_accuracy: 0.9583\n"
     ]
    }
   ],
   "source": [
    "history = model.fit_generator(\n",
    "    train_data, \n",
    "    epochs=50,\n",
    "    steps_per_epoch=steps_per_epoch,\n",
    "    validation_data=validation_data,\n",
    "    validation_steps=6\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_history(history):\n",
    "    fig, ax = plt.subplots(1, 2, figsize=(15,5))\n",
    "    ax[0].set_title('Loss')\n",
    "    ax[0].plot(history.epoch, history.history[\"loss\"], label=\"Train loss\")\n",
    "    ax[0].plot(history.epoch, history.history[\"val_loss\"], label=\"Validation loss\")\n",
    "    ax[1].set_title('Accuracy')\n",
    "    ax[1].plot(history.epoch, history.history[\"accuracy\"], label=\"Train acc\")\n",
    "    ax[1].plot(history.epoch, history.history[\"val_accuracy\"], label=\"Validation acc\")\n",
    "    ax[0].legend()\n",
    "    ax[1].legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3AAAAE/CAYAAAAHeyFHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nOy9eZhcZZm/fz/pNd1JZxeysISAmq2z0LKILEFE0JElipIBQVwQhVHhy4w4iiijP5wR1xGXqLggGhkcHVQQFUFgRpFAFiQIhAQk6RA6W+97P78/3nOqT1dqOaf2qn7u6+qrqk6d5a2l67yf83kWUVUMwzAMwzAMwzCM0mdCsQdgGIZhGIZhGIZhhMMEnGEYhmEYhmEYRplgAs4wDMMwDMMwDKNMMAFnGIZhGIZhGIZRJpiAMwzDMAzDMAzDKBNMwBmGYRiGYRiGYZQJJuAMwzAMwzAMwzDKBBNwhpEHROR5ETmj2OMwDMMwjHwjIg+IyH4RqSv2WAxjPGACzjAMwzAMw8gIETkSOBlQ4JwCHre6UMcyjFLDBJxhFBAReZ+IbBWRfSJyl4jM8ZaLiHxJRF4WkXYR2SwiS7zn3iQiW0SkU0R2isi1xX0VhmEYhhHjEuDPwPeBS/2FIjJRRL4gIi9457WHRWSi99zrROT/ROSAiLwoIu/ylj8gIu8N7ONdIvJw4LGKyJUi8izwrLfsK94+OkTkMRE5ObB+lYj8q4g8551DHxORw0TkFhH5QvBFiMgvReQj+XiDDCPXmIAzjAIhIqcDNwFvB2YDLwDrvKfPBE4BXglMBd4B7PWe+y7wflWdDCwB/lDAYRuGYRhGKi4Bbvf+3igih3jLbwaOBV4LTAf+BRgRkcOBe4D/BGYBy4GNEY53HnA8sMh7/Ki3j+nAj4H/EpF677lrgDXAm4Am4N1AD/ADYI2ITAAQkZnA64GfRHnhhlEsTMAZRuG4CLhVVR9X1X7gY8CJXvjJIDAZeDUgqvqUqu7ythsEFolIk6ruV9XHizB2wzAMwxiDiLwOOAK4Q1UfA54D/tETRu8GPqyqO1V1WFX/zzv3XQT8XlV/oqqDqrpXVaMIuJtUdZ+q9gKo6o+8fQyp6heAOuBV3rrvBT6hqk+rY5O37l+AdpxoA7gQeEBVd2f5lhhGQTABZxiFYw7OdQNAVbtwLttcVf0D8DXgFmC3iKwVkSZv1bfirh6+ICJ/FJETCzxuwzAMw0jEpcBvVXWP9/jH3rKZQD1O0MVzWJLlYXkx+EBE/p+IPOWFaR4ApnjHT3esHwAXe/cvBm7LYkyGUVBMwBlG4WjFXakEQEQagRnATgBV/aqqHgssxoVS/rO3/FFVPRd4BfAL4I4Cj9swDMMwxuDls70dOFVEXhKRl4CrgWW4NIE+YEGCTV9MshygG2gIPD40wToaGMPJwEe9cUxT1ak4Z01CHOtHwLkisgxYiDu/GkZZYALOMPJHjYjU+3844XWZiCz3Si3/f8Ajqvq8iLxGRI4XkRrcCawPGBaRWhG5SESmqOog0AEMF+0VGYZhGIbjPNz5aBEuB205Tgg9hMuLuxX4oojM8YqJnOid+24HzhCRt4tItYjMEJHl3j43AqtFpEFEjgbek2YMk4EhoA2oFpFP4nLdfL4D/JuIHOMVC2sWkRkAqroDlz93G/AzPyTTMMoBE3CGkT/uBnoDfycD1wM/A3bhrgpe6K3bBHwb2I8Ls9yLSwAHeCfwvIh0AFcwGvJhGIZhGMXiUuB7qvp3VX3J/8OlA1wEXAc8gRNJ+4B/Byao6t9xaQH/z1u+EefaAXwJGAB240Icb08zhntxBVGewZ07+xgbYvlF3MXT3+IugH4XmBh4/gfAUix80igzRFXTr2UYhmEYhmEYFYSInIILpTxSVUeKPR7DCIs5cIZhGIZhGMa4wktZ+DDwHRNvRrlhAs4wDMMwDMMYN4jIQuAArtjKl4s8HMOIjIVQGoZhGIZhGIZhlAnmwBmGYRiGYRiGYZQJJuAMwzAMwzAMwzDKhOpiDyCemTNn6pFHHlnsYRiGYRgF4LHHHtujqrOKPY5ywc6RhmEY44NU58eSE3BHHnkk69evL/YwDMMwjAIgIi8UewzlhJ0jDcMwxgepzo8WQmkYhmEYhmEYhlEmmIAzDMMwjDSIyFki8rSIbBWR6xI8f4SI3Ccim0XkARGZF3juP0TkSRF5SkS+KiLiLT9WRJ7w9hlbbhiGYRipMAFnGIZhGCkQkSrgFuBsYBGwRkQWxa12M/BDVW0GbgRu8rZ9LXAS0AwsAV4DnOpt8w3gcuAY7++s/L4SwzAMoxIouRy4RAwODrJjxw76+vqKPRQjBPX19cybN4+amppiD8UwDCMXHAdsVdVtACKyDjgX2BJYZxFwtXf/fuAX3n0F6oFaQIAaYLeIzAaaVPVP3j5/CJwH3JPfl2IYxnjC5tClTybz5rIQcDt27GDy5MkceeSRWIRJaaOq7N27lx07djB//vxiD8cwDCMXzAVeDDzeARwft84m4K3AV4DzgckiMkNV/yQi9wO7cALua6r6lIi0ePsJ7nNuvl6AYRjjE5tDlzaZzpvLIoSyr6+PGTNm2BevDBARZsyYYVd6DMOoJBKdfDTu8bXAqSKyARciuRMYEpGjgYXAPJxAO11ETgm5T3dwkctFZL2IrG9ra8v0NRiGMQ6xOXRpk+m8uSwEHGBfvDLCPivDMCqMHcBhgcfzgNbgCqraqqqrVXUF8HFvWTvOjfuzqnapahcuRPIEb5/zUu0zsO+1qtqiqi2zZlnLPMMwomHzstImk8+nbARcMdm7dy/Lly9n+fLlHHroocydOzf2eGBgINQ+LrvsMp5++unQx/zOd77DRz7ykUyHbBiGYeSOR4FjRGS+iNQCFwJ3BVcQkZki4p9TPwbc6t3/O86ZqxaRGpw795Sq7gI6ReQEr/rkJcD/FOLFGIZhFIpizKHHA2WRA1dsZsyYwcaNGwH41Kc+xaRJk7j22mvHrKOqqCoTJiTWxN/73vfyPk7DMAwj96jqkIhcBdwLVAG3quqTInIjsF5V7wJOA24SEQUeBK70Nr8TOB14Ahci+RtV/aX33AeA7wMTcc6cFTAxDKOisDl0fjAHLgu2bt3KkiVLuOKKK1i5ciW7du3i8ssvp6WlhcWLF3PjjTfG1n3d617Hxo0bGRoaYurUqVx33XUsW7aME088kZdffjnlcbZv386qVatobm7mDW94Azt2uLz3devWsWTJEpYtW8aqVasAeOKJJ3jNa17D8uXLaW5uZtu2bfl7AwzDiE73XmjdUOxRGBFR1btV9ZWqukBVP+st+6Qn3lDVO1X1GG+d96pqv7d8WFXfr6oLVXWRql4T2Od6VV3i7fMqVU2YA2cYhlFp5HMO/ec//5kTTzyRFStWcNJJJ/Hss88CMDQ0xNVXX82SJUtobm7m61//Ogz18cj/PsSJJ57IsmXLOP744+np6SnY+5ApJuCyZMuWLbznPe9hw4YNzJ07l8997nOsX7+eTZs28bvf/Y4tW7YctE17ezunnnoqmzZt4sQTT+TWW29NsOdRPvjBD/Le976XzZs3c8EFF8RCKz/96U9z3333sWnTJn7+858D8PWvf51rr72WjRs38uijjzJnzpzcv2jDMDLnT/8Jt51f7FEYhmEYRlHJ1xx64cKFPPzww2zYsIHrr7+eT3ziEwB84xvfoLW1lU2bNrF582YuvPBC+l7ayoX/+I/ccsstbNq0id/+9rfU1dXl/bVnS9mFUH76l0+ypbUjp/tcNKeJG96yOKNtFyxYwGte85rY45/85Cd897vfZWhoiNbWVrZs2cKiRWP7vU6cOJGzzz4bgGOPPZaHHnoo5TEeeeQRfvWrXwFwySWXcP311wNw0kkncckll3DBBRewevVqAF772tfymc98hhdeeIHVq1dz9NFHZ/S6DMPIE70H3J8qWGK5YRiGUSDGyxz6wIEDXHLJJTz33HNjlv/+97/nIx/5CFVVVQBMnz6dDRvu5/C5s1m5ciUAU6ZMyei1FBpz4LKksbExdv/ZZ5/lK1/5Cn/4wx/YvHkzZ511VsKyoLW1tbH7VVVVDA0NZXTsb3/723z605/m+eefZ9myZezfv593vvOd/PznP6euro43vOENPPjggxnt2zCMPDHYCygMh0veNgzDMIxKJF9z6I9//OO88Y1v5K9//Su/+MUvYvtR1YMqPurISFleSy07By5TlV8IOjo6mDx5Mk1NTezatYt7772Xs846K+v9nnDCCdxxxx2sWbOGH/3oR5xyyikAbNu2jRNOOIHjjz+eu+66i507d7J//36OPvpoPvzhD/Pss8+yefPm2PqGYZQAgz2jt9WlH6ZhGIZhVAbjZQ7d3t7O3LlzAfj+978fW37mmWfyjW98g5NPPpmqqir27dvH4mPm88KLrTz++OOsXLmSjo4OGhsbYy5dqRLKgRORs0TkaRHZKiLXJXj+ChF5QkQ2isjDIrLIW36kiPR6yzeKyDdz/QJKiZUrV7Jo0SKWLFnC+973Pk466aSc7PdrX/saa9eupbm5mZ/+9Kd86UtfAuDqq69m6dKlLF26lDPOOIMlS5bw4x//mMWLF7N8+XK2bdvGxRdfnJMxGIaRIwZ7vVtrdm8YhmEYkNs59Ec/+lH++Z//+aB9vP/97+fQQw+lubmZZcuWcccdd1BXW81Pvvl5PvCBD7Bs2TLOPPNM+vv7s305eUfSFb0SkSrgGeANuMajjwJrVHVLYJ0mVe3w7p8DfFBVzxKRI4FfqeqSsANqaWnR9evXj1n21FNPsXDhwrC7MEoA+8wMIwnfezO88DD80+MwY0GxR1N0ROQxVW0p9jjKhUTnSMMwjGTYfCwFOgK7NsGEGjg0tFTJC4k+p1TnxzAO3HHAVlXdpqoDwDrg3OAKvnjzaMT1ujEMwzDiGex2t0PmwBmGYRhG0dAR/05Rh5EJYQTcXODFwOMd3rIxiMiVIvIc8B/AhwJPzReRDSLyRxE5OavRGoZhlDuxEMre4o7DMAzDMMYzI56Aiwm58iGMgEtUm+Ugqaqqt6jqAuCjwCe8xbuAw1V1BXAN8GMRaTroACKXi8h6EVnf1tYWfvSGYRjlRqyIiQk4wzAMwygavnBLk05WioQRcDuAwwKP5wGtKdZfB5wHoKr9qrrXu/8Y8BzwyvgNVHWtqraoasusWbPCjt0wDKP8MAfOMAzDMIqPDvt3yk7EhRFwjwLHiMh8EakFLgTuCq4gIscEHr4ZeNZbPssrgoKIHAUcA2zLxcANwzDKEl+4DZmAMwzDMIyiEQydLLMwyrR94FR1SESuAu4FqoBbVfVJEbkRWK+qdwFXicgZwCCwH7jU2/wU4EYRGQKGgStUdV8+XohhGEbJo2ohlIZhGIZRCowEBVzlOXCo6t2q+kpVXaCqn/WWfdITb6jqh1V1saouV9VVqvqkt/xn3vJlqrpSVX+Zv5eSP0477TTuvffeMcu+/OUv88EPfjDldpMmTQKgtbWVt73tbUn3na4k9Je//GV6enpij9/0pjdx4MCBMENPyac+9SluvvnmrPdjGEZIhgdGr/KZgDMMwzAqnJKeQ8dCKAHKy4ELJeDGO2vWrGHdunVjlq1bt441a9aE2n7OnDnceeedGR8//st39913M3Xq1Iz3ZxhGkRjsCdw3AWcYhmFUNiU9h9YKd+DGO29729v41a9+FevM/vzzz9Pa2srrXvc6urq6eP3rX8/KlStZunQp//M//3PQ9s8//zxLlrgGgb29vVx44YU0Nzfzjne8g97e0UncBz7wAVpaWli8eDE33HADAF/96ldpbW1l1apVrFq1CoAjjzySPXv2APDFL36RJUuWsGTJEr785S/Hjrdw4ULe9773sXjxYs4888wxx0nExo0bOeGEE2hubub8889n//79seMvWrSI5uZmLrzwQgD++Mc/snz5cpYvX86KFSvo7OzM+L01jHFFULRZDpxhGIZR4ZT0HPqrX2fJ6Rew5PQL+PJXvho7Xpg59C9/+UuOP/54VqxYwRlnnMHu3bsB6Orq4rLLLmPp0qU0Nzfzs5/9DIDf/OY3rFy5kmXLlvH6178++zdWVUvq79hjj9V4tmzZctCyQvOmN71Jf/GLX6iq6k033aTXXnutqqoODg5qe3u7qqq2tbXpggULdGRkRFVVGxsbVVV1+/btunjxYlVV/cIXvqCXXXaZqqpu2rRJq6qq9NFHH1VV1b1796qq6tDQkJ566qm6adMmVVU94ogjtK2tLTYW//H69et1yZIl2tXVpZ2dnbpo0SJ9/PHHdfv27VpVVaUbNmxQVdULLrhAb7vttoNe0w033KCf//znVVV16dKl+sADD6iq6vXXX68f/vCHVVV19uzZ2tfXp6qq+/fvV1XVf/iHf9CHH35YVVU7Ozt1cHDwoH2XwmdmGCXHnq2qNzS5v/v+rdijKQlwudRFP/eUy1+ic6RhGEYySmE+VrJz6EWv1q5n/1c7n3lYFy1cGGkOvW/fvthYv/3tb+s111yjqqr/8i//EptD++u9/PLLOm/ePN22bduYsQZJ9DmlOj+mLWJSctxzHbz0RG73eehSOPtzKVfxLeBzzz2XdevWceuttwJOAP/rv/4rDz74IBMmTGDnzp3s3r2bQw89NOF+HnzwQT70IdfnvLm5mebm5thzd9xxB2vXrmVoaIhdu3axZcuWMc/H8/DDD3P++efT2NgIwOrVq3nooYc455xzmD9/PsuXLwfg2GOP5fnnn0+6n/b2dg4cOMCpp54KwKWXXsoFF1wQG+NFF13Eeeedx3nnnQfASSedxDXXXMNFF13E6tWrmTdvXsr3zjAMDwuhNIzsGR6CH54Dp1wLC04v9mjgno9C3WQ4/RPp1zUMgMdvg7/9Gtb8BCRRu+U8YXPoGA8//DDnv/mNNDZMBGD1eW+JNIfesWMH73jHO9i1axcDAwPMnz8fgN///vdjQkanTZvGL3/5S0455ZTYOtOnT0/5foXBQihDct5553Hffffx+OOP09vby8qVKwG4/fbbaWtr47HHHmPjxo0ccsgh9PX1pdyXJPhn3b59OzfffDP33Xcfmzdv5s1vfnPa/WiKeN26urrY/aqqKoaGhlLuKxm//vWvufLKK3nsscc49thjGRoa4rrrruM73/kOvb29nHDCCfztb3/LaN+GMe4IijYTcIaRGT174IX/hb/+d7FHAvufh0e+Bc/dX+yRGOXEtvvhmXtg2wPFHklBKIs5dOBxmDn0P/3TP3HVVVfxxBNP8K1vfSt2PFU9aIyJlmVL+TlwaVR+vpg0aRKnnXYa7373u8ckXra3t/OKV7yCmpoa7r//fl544YWU+znllFO4/fbbWbVqFX/961/ZvHkzAB0dHTQ2NjJlyhR2797NPffcw2mnnQbA5MmT6ezsZObMmQft613vehfXXXcdqsrPf/5zbrvttsivbcqUKUybNo2HHnqIk08+mdtuu41TTz2VkZERXnzxRVatWsXrXvc6fvzjH9PV1cXevXtZunQpS5cu5U9/+hN/+9vfePWrXx35uIYx7gg6cEOpTy6GYSShr93d7nysuOMA+Mu3AR0dk2GEobvN3T7yLViwqnDHtTn0mH29653f5Lr3vx1V+Pldv+K2H90e+jW1t7czd+5cAH7wgx/Elp955pl87Wtfi9Wl2L9/PyeeeCJXXnkl27dvZ/78+ezbty9rF678BFwRWbNmDatXrx5jjV500UW85S1voaWlheXLl6cVMh/4wAe47LLLaG5uZvny5Rx33HEALFu2jBUrVrB48WKOOuooTjrppNg2l19+OWeffTazZ8/m/vtHr/KtXLmSd73rXbF9vPe972XFihUpwyWT8YMf/IArrriCnp4ejjrqKL73ve8xPDzMxRdfTHt7O6rK1VdfzdSpU7n++uu5//77qaqqYtGiRZx99tmRj2cY45IxDlxP8vUMw0hOX4e7ffkp6O904YvFYKAbNngXTfs7ijMGozzpdkU0eOY3sG87TJ9f3PEUgJKcQ695K8e9+RIA3vue90SaQ3/qU5/iggsuYO7cuZxwwgls374dgE984hNceeWVLFmyhKqqKm644QZWr17N2rVrWb16NSMjI7ziFa/gd7/7XajjJENSheEVg5aWFo3v6fDUU0+xcOHCIo3IyAT7zAwjAX/9Gdz5bqiuh6NWwT+uS79NhSMij6lqS7HHUS4kOkeOO579Pdz+Vnf/0l/C/FOKM471t8KvrobDjoddm+ATu4szDqP8+PzRMGclPHcfHH8FvPGzeTuUzcdSsOdZd2FVh2HqEdCQfW5apiT6nFKdHy0HzjAMo1D4DtzE6ebAGUam9B0Yvb+jSGJWFR5ZC4c2wzFnupDoof7ijMUoL0ZGoGcvzG6Ghee4gib9XcUe1fhER2CCH4xYWoZWOkzAGYZhFApfwDXMsBw4w8gUP1yxrql4eXDbH4S2p5x7Uj/FLeuzMEojBL37nXBonOW+P/3tsPmnxR7V+CQo4IJNvcsAE3CGYRiFwnfdGqaZA1dmiMhZIvK0iGwVkesSPH+EiNwnIptF5AERmectXyUiGwN/fSJynvfc90Vke+C55YV+XWWJXzBk/inFE3CPfMtdiFnyVqifOnZchpEKv4BJ40w47DiYvdx9n0ospWlcMEbAldf7XzYCrtRy9Yzk2GdlGEkYE0JpDly5ICJVwC3A2cAiYI2ILIpb7Wbgh6raDNwI3ASgqver6nJVXQ6cDvQAvw1s98/+86q6Md+vpSLoa3eTriNPhs5d0L6zsMff/zw8fTcc+y6oqYf6ptFxGUY6YgJulusBd/z7Yc/TeW0pYPOyJIwMQ1WVu19EBy6Tz6csBFx9fT179+61L2AZoKrs3buX+vr6Yg/FMEqPwR5XwKS20frAlRfHAVtVdZuqDgDrgHPj1lkE3Ofdvz/B8wBvA+5RVbNfs6Gvw4UtzvNy+3cWOA/u0e+ATICW97jHfghlvwk4IwRBAQeweDU0zIS/rM3L4WwOnQIdASmuA5fpvLks2gjMmzePHTt20NbWVuyhGCGor69n3rx5xR6GYZQeg71QM9H9DZmAKyPmAi8GHu8Ajo9bZxPwVuArwPnAZBGZoap7A+tcCHwxbrvPisgnceLvOlW1Shjp6Gt3ounQpVBV6wqZLEqkl/PAQDc8/kNYdA5McT2gRnPgTMAZIfBbCPgCrqYeWi6DB2/OS0sBm0MnQRXad0N9n7soVNcHEw+k3y4PZDJvLgsBV1NTw/z5ld8jwzCMCmewB2oanAtnDlw5IQmWxV+uvRb4moi8C3gQ2AkMxXYgMhtYCtwb2OZjwEtALbAW+Cgu/HLswUUuBy4HOPzwwzN9DZVDX7srYFJd50RcIfPgNv/UHf+4948uq7MQSiMC3W3OwZ04bXRZy7vh4S85dzfHLQVsDp2E3v3w7yfCG2+CB26CFRfDWTcVe1ShKYsQSsMwjIog5sA1eL1nLKSlTNgBHBZ4PA9oDa6gqq2qulpVVwAf95YFZ/RvB36uqoOBbXapox/4Hi5U8yBUda2qtqhqy6xZs3LzisqZ/o5R12tuC7RugOGh1NvkgmDrgMNPGF1uVSiNKHS3uQI4E6pGlzXNsZYChWag293WNrqLQWVWGdoEnGEYRqEY7HXiraYeUOsbVT48ChwjIvNFpBYXCnlXcAURmSki/jn1Y8CtcftYA/wkbpvZ3q0A5wF/zcPYKw8/hBJcHtxgjyvpn2+CrQMkYMrWNoJUmQNnhKO7zeW8xWMtBQqLL+DqJkFVXdmdj03AGYZhFAo/hLKmwT22PLiyQFWHgKtw4Y9PAXeo6pMicqOInOOtdhrwtIg8AxwCxOKgRORInIP3x7hd3y4iTwBPADOBz+TxZVQOfe2jlR/nHutuCxFGGWwdEETEjccEnBGG7j2uhUA8fkuBv6y16IxCMOA5nbWTytKBK4scOMMwjIrAd+Cq60cfB/MgjJJFVe8G7o5b9snA/TuBO5Ns+zyuEEr88tNzO8pxQl/HaO+16Ue5/6Ed611Z/3zhtw44+RrPQY+jfspog3HDSEV3G8xedvByEefC/eIK2P5HOOq0Qo9sfOGHqtY2unNymTlwJuAMwzAKxWCPC53xHTgrZGIY0RgehMHu0RBKEefC5duBi28dEE/9lMp14IaH4G+/cpU+JVE9nySouu2OeSNU1+ZvfMWk7Rnn3MxuDr9N957RCpTxLFkNv/0E/OEz8NITideZMg8Wnx99rMnI9PMtd7LNgdu3zRVC8aMACowJOMMwjEIx0OMVMQk4cIZhhMcvFOJXfgRXyGTrv0N/J9RNzs9xn/09LFg12jognroKDqF85h74r0vh8j/CnOXht9vzDPz0YnjHj2DhW/I3vmJy779C12644qFw6w/1uzy3ZAKuug5O/CDcdyPseDT5fmYc7Sqw5oJnfuM+3/c/mNgZrFRiAm5SZg7c/TdB6+PwTwWsghvABJxhGEahCFahhLKLuTeMouM3y/YdOPAaequrRjn/lPwct+/AaOPwRNRPcT28KpE9z7rb/s5o2/liu5Krc3a+BB07wq8f6wGXIAfO5+T/57WpSJAH198J/3msy8c892uRhpqUdq/F5XirfhmfA+cLurD0d0BX8XrrWRETwzCMQhHsA+c/NgwjPH0JBJwfwrRjfX6PGzxmPPVTK9eB2/ecu416wckv0lTJv3PdbS6MLmw0Rbc34U/mwPnUTXJucvxf0xxofjs88V/Qsy+7sft0eB1RxtsFxWxDKAd73QWlQrQwSYAJOMMwjEIR78ANjrMTpmFkS0zABUIoG6a7Yib5yoMbHnQixC+ckohKrkK5d5u7jRry7f++RXU2yoWREejxHLWO1tTr+sQcuCz6OR73fic2Hv9B5vsIYgLOE3ARQyj996tI//cm4AzDMArByDAM9wf6wFHZV6YNIx/44Xjxblg+C5nEjtmUfJ36KTDQ6f7PK419mQq4nrG3lUbfARjx3JfOXeG2iTlwKUIo03HIIhcq/Jfv5Mb98cc+7gRcJ1RPdA3Vq+szcOC873Xv/tyPLQQm4AzDMAqBP/mpmej+YPydMA0jWxKFUIIrZNK5C9p35uGYBxIfM4j/XKW1Eujvgq6X3GNRDEYAACAASURBVP2ofSuHKtyB8900CO/A9eTAgQPnwnXsgKd/nd1+ADq8/5nxFhEy0O3cN8jMgfPfr94chbJGxAScYRhGIQgKuGpPwFXqlWnDyBe+gKuLc8P8AiM785AHl0w0BvHHU2lhlPsDhVmiTvAHKzwHrjtQwCJ0CGUbVNVlXy31VWfDlMPhkbXZ7UcVOsarA9ftcg0hMwfOX7+UHTgROUtEnhaRrSJyXYLnrxCRJ0Rko4g8LCKLAs99zNvuaRF5Yy4HbxiGUTb4k5iahlEHbrxd8TSMbOnvAORgAXfoUqiqzU8hk/4ErQvi8cVdpVVc3Pvc6P2MHbhxIOBCh1B6PeCy7bc2oQqOex+88HDyfnFh6N3vQvthfAq4Wl/A1cHwQLTt/QsUpSrgRKQKuAU4G1gErAkKNI8fq+pSVV0O/AfwRW/bRcCFwGLgLODr3v4MwzDGF4lCKCv1yrRh5Iu+diekJsRNX6rrnIjLRx5cGAcuJuAqzIHz898gCweuUkMoPQE3cdpoGGKYbbLJfwuy8p3uguAj38p8H8FxjzsB1zUaQlnlVaHUBK0bklEGDtxxwFZV3aaqA8A64NzgCqoavOTUyGjzinOBdarar6rbga3e/gzDMMYXQQeuqhZkwvg7YRpGtvS1Jy8mMrfF9YLLdVnvUAKuQkMo9z0Hkw7xQszMgRuDnwN3yJLRMMS027Rln//mM3Fa9i0FguMebxEh/QEBV10POjJalCYM/gWKXLVziEgYATcXeDHweIe3bAwicqWIPIdz4D4UZVvDMIyKJ+jAibg8uKhV3QxjvNPXkVxIzT3WXShp+1vujwnpq1BC5RUx2bsNpi9wE1yrQjmW7jaYOB2mHhGtjUCuBBxk31JgXDtwcUVMIPx7MDwI6lWcLWEHLlGg7kEeo6reoqoLgI8Cn4iyrYhcLiLrRWR9W1vxupobhmHkjZiA83rA1ZiAM4zIpGqona9CJn3tgEBtisITlRxCOf2ozH6vKr0PnO+mNc2Grt3pnV/V3IZQQvYtBTp3EftuR63CWO6MyYHzWvuEfQ+CFyVKWMDtAA4LPJ4HpLrUsA44L8q2qrpWVVtUtWXWrBxemTAMwygVYiGUE0dvTcAZRjT8HLhETD/KhZXlupCJH7YZn3cXpBKrUPotBGZ4Ai5ylb5Kr0LpuWmTZzs3pvvl1OsPdLn3MJcCDgItBe6Ovm3HTpj0CleNMWqIbLkz0DW2iAmE/44Hw01LuI3Ao8AxIjJfRGpxRUnuCq4gIscEHr4ZeNa7fxdwoYjUich84BjgL9kP2zAMo8wIhlD6t+PthGkY2dKfwoETyU9D7/4OqEuR/wauKmDt5MqqQum3EJh+VGYh34MVngPXs8eJsSYvMyhdHlysiXeOjYpYS4EMipl07IKmOZn1QSt3xoRQRnTggufuUnXgVHUIuAq4F3gKuENVnxSRG0XkHG+1q0TkSRHZCFwDXOpt+yRwB7AF+A1wpaofNGoYhjGOCBYxgcxySgxjvJMqhBJcIZOXn4L+zsId06d+SmU5cH4LgekLoCaLPlmVXIXSD6EE6EyTB9edoybe8WTTUqBzF0yeM/7OR8ODrn1CzIGrdbehQyi97/aEmtIVcACqereqvlJVF6jqZ71ln1TVu7z7H1bVxaq6XFVXecLN3/az3navUtV78vMyDMMwSpyDHLiG8XXCNIxsGRnxipikKCYyrwVQV40yV4QWcE3QdyB3xy02fguB6fM9By7DNgKV6MAND7qJ+xgHLp2A8x24HIdQAqy42H1GUV24jp1OgFbXjy8HbqDL3R7kwIX8jvsO3OTZpS3gDMMwjCyJd+BqxtkVT8PIloEuQNM4cMe62xf+lLvjphONPvVTKqsKpd9CoG6y93sVUYj5v28jg07wVBI9e91t40xomOFaw4QWcHmo9dAwHZa9w7UU6O8Kt81At7s40TQnszYRqejrgP9sgRf+L3f7zCV+YZ2DqlBGdOCaZrv3MNetS0JgAs4wDKMQDPaCVEFVjXtc02A5cIYRhTD92BqmwxGvgw0/gpEcZWyM2xBKr4UAeBP8DIuYQOVVogyKMRGYfGh4AdeQBwcOYMHr3We0d2u49f2cvclzvBDZHDpwe591f3/7de72mUv872NdfBXKiA5c0xx3W4T/exNwhmEYhWCw14k28bqrjLecA8PIFn+SlKwKpc/x74f2v8PTOcraCCvg6poqS8D5LQQguzYCUHmVKOPdtKa5Xkn+VNvscd+Rmvr8jGmGJ7b3PRdufT9nrykPOXC+OMx1RdhcEQuhjK9CGdaB80MoPQFXhDBKE3CGYRiFYLBnNP8NvBy4cdY41TCywQ9PTCemXvUmaJoHj3wz+2OOjHhVKEOGUFZKFcpgCwHIvI2AeNPMSsuDiy9IMnn22KbYCbfJcQ+4eKbNd7d+7mI6fJEVC6HMoQPnu5G7NpZm+OxBIZSeqB6OKOD8AjZFaCVgAs4wDKMQDPbGCbgMckoMYzwTJoQSoKoajnsvPP8Q7H4y9brpGOgkbd6djx9CqZrdMUuBYAsByLyNwMTp3v1KDaH0BFnTHCeIUn32ftXKfFHb4ByhvWEFnCc4J8/OLEQ2Fb67N9SX/f9gPuiPK2JSFdGB89+rJnPgDMMwKpvBntECJpD7E6ZhVDphBRzAykvd/9hf1hbumPVNrqFzJeR7BVsIQIZtBHpdgQ+oQAeuDSZUQ/1U97hpjnu9qaqQ+o2/88mMBRFCKHe5/oZ1k7w+cDk8H3W0jrpaO0swjDLmwGXayNt34LwKpCbgDMMwKpTBXneF1KemwYm6SrhaPw4QkbNE5GkR2Soi1yV4/ggRuU9ENovIAyIyz1u+SkQ2Bv76ROQ877n5IvKIiDwrIj8VkdpCv66yoi9kCCW4YiZLL4BNP4WeLMKbYscMGUIJlVGJMthCAJwDN9TnQkrDMtg3KuAq0YHzC5iAc7EgdSGT7j35DaEE93mFDqFsHXWQMgmRTbfvQ5tdwZYdj+Vuv7kiaRuBiA6c/7mbgDMMw6hQ/CImPjURTxhG0RCRKuAW4GxgEbBGRBbFrXYz8ENVbQZuBG4CUNX7vR6py4HTgR7gt942/w58SVWPAfYD78n7iylnwhYx8Tn+/c4V2XBb9scMG0IZ3KacCbYQgMDvVchJ/siwyydqrFQHLk6MxXrBJSlkMjICPQVw4KYvcOIyTC5mR+toDld1XW5zsjt3OXE4rwV2lqKAy9aB8wXcoYBkd5EoQ0zAGYZhFIJERUz85UapcxywVVW3qeoAsA44N26dRcB93v37EzwP8DbgHlXtERHBCbo7ved+AJyX85FXEn0H3P9NdUij8tClrqXAX76TeUuBKALOF5aVIOCCLQTAOXAQocy6t17Mgauw37n4fDZfCCUrZNK7H3SkAALOy1kM48J1tI5WUaye6AR3LiJCVEfdvbktsOeZ0vufSFbEJEobAalywq9+ijlwhmEYFUt8EZOoJwyjmMwFXgw83uEtC7IJeKt3/3xgsojMiFvnQuAn3v0ZwAFV9TvAJtonACJyuYisF5H1bW1tGb6ECiBsNcggx1+eXUsBPxwyVBVKLx+qEipRBlsIwOhvV9hCJoNxAq4S8gKDxAu4SZ4Tk6yVQHzRk3wRtpXA8BB0vzwaQhnVgUpFX7sT7E1zYN6xgMLOx7Pfby4Z6HSidUKVe1xVAwgMDYTbfrBv9H+iYboJOMMwjIplsDsuhNJ34KwXXBkgCZbFX6q+FjhVRDYApwI7AV+cISKzgaXAvRH26RaqrlXVFlVtmTUrz1fwS5mw/diCvOrNrqXAX76V+TFhVJylolJCKONbCMDoZDVqo+OKdeDiwiGra93jZA5cfN+4fBG2lUDXbucI+s5h1M83FX4e4OTZMGelu19qhUwGukfdN3C5jFEKiw32jF6EnTjNBJxhGEbFkqiNgL/cKHV2AIcFHs8DxlQrUNVWVV2tqiuAj3vLgjP5twM/V1W/KdIeYKqIVCfbpxFHJgLObymw/UHYvSWzY0LIIiZ+CGWKSoTlQHwLARidrEZ14Pw2ApWUAzfQ7Sbw8W6a30ogEb6Aa8izAxe2lYAvsvzcvaiNrFPRGdj3xKkw85WlV8gkXsCBE+FRipj4F2EnTrM+cIZhGBXLQUVMvPsWQlkOPAoc41WNrMWFQt4VXEFEZor4XYv5GHBr3D7WMBo+iaoqLlfubd6iS4H/ycPYK4e+jnBCKp5YS4EMXLi+dve/WlWTfl0/zLLcq1DGtxCA6CGUvgNXN8n12KqkKpTxTbx9muakCKFMsk0+mH5U+hDKzoBLBqM5jrm4oBgTh96+5x7rHLhSqrg80D1aoMcnkgPXO3oR1hw4wzCMCkX14CImsSvaFXRlukLx8tSuwoU/PgXcoapPisiNInKOt9ppwNMi8gxwCPBZf3sRORLn4P0xbtcfBa4Rka24nLjv5vFllD+ZOHAwtqVA1IlWlGPW1Lv/63IPoYxvIQCBnN2wDpy3XvVE5wpVkgOXTIxNnp0mhFLcdzHfzDgqfQhlPh0434WcHBBw3W1w4O/Z7ztXDHQlcODqojlwsRDK4uTAVadfxTAMw8iK4QGXb5CwCqU5cOWAqt4N3B237JOB+3cyWlEyftvnSVCgRFW34SpcGmHIVMCBaymw4TZ4/DY46UP5O2ZdUwUIuLgWAhBw4CI2Oq6ph5rGyrpQlawgSdMcN5GPD5f3t2mYMVo0I58EWwkkc6w7Wp0z6gvKqAI9FZ2tLlTUF4XzWtztzsdg2hHZ7z8X9Hcd/N5EduC8z3jiNPc/PzzkQrYLhDlwhmEY+cafvCTqA1dJExvDyBeqmVWh9PFbCjz67WgtBfo7ogm4+inlX4UyvoUABIpchA2h9CbCMQeukkIokxQk8Ss6JmrmHV+1Mp+EaSXQucv1MPMbkeeyL2mwvxzAIUucOCqlfnAJc+Dq3MXWMIxx4Ka52wJfuDEHzjAMI9/ErkYHHbgcVv0qNJ274RdXwGs/BAtWFXs0xnhgqM9NrjJ14MC1FLjjEtdSYOE/hNumrz1a4Yn6KbmZyO15Fp64kySFSV1hiKVvS/xctuzbBkefMXZZLOQ7EweuofQvVL3wf+6zO2Rx+nWTFSTxBVznrtFy/rFt9uS/hYBPsJXAnOWJ1+loHQ2fhOhFalLRsQumBPZdVQOzl8GOEqpEOdA92sTbJ6oD5ws338Xs3T/auL4AmIAzDMPIN7HJTMCBiyWNl/jEJhG9++G5P8Dyi4o9EmO8EKWhdjJe9Waom+K+u1EEXLwblYr6HIVQ/u9XXMhnMmQCLDo3XHGVKCRqIQBZOHD1zuko9Ry4X13tBM07/zv9ut17XFhobcPY5ZPTOHCzl2U/zjCEaSXQ0QpzV44+rs6lA7dzNGzSZ24LrP8uDA/m/jubCQNdBwu4qgg5cIO9BztwBc6DMwFnGIaRb2IhlAkcuHLMgfMrysWHoBhGvvDDErMRcFXVLrSrO0Iz9L4MQijbd0QfWzwdO13xh/f94eDn1n8PfvUR18tryrzsjxUkUQsByKCNQOCiVU0D9OzJzfjyRe/+8O5Ld1tiN80PG0wo4PYULoQyXSsBVTfG4EWMXOXADfa5kvpNcSm/c1fCn2+B3U8mdwULSbIQyrAtQIb6xubAQcFbCVgOnGEYRr5JFUJZjg7cQIKcPsPIJ7lw4MBNortDiglVr4hJhLy7XIVQdrSOVvGLJ1WuVbYkaiEA0dsIBEMoy6EKZV+7q5I4FCIHKlk+W91kl6MZ30pgqB/62wsn4CB1K4He/TDcP+oYQu5y4PzX3hT33Y0VMimBMMrhQff6E4ZQRnDgDhJwhXXgTMAZhmHkm0RFTKpqXRhUOebA+a/HHDijUORMwM0M78AN9sLIYHGqUHbsOtjF8MmngIu1EEjiwIX9vRoKtBEo9SqUQ/3udekIHHgh/fqp3LRErQRibQcKlAMHqVsJ+ONrCgi4qJ9vMjri+sv5TD3C5QzufDy7/eeCgS53m7CNQNjvd99oGoQJOMMwjAolkQMn4k4AuUgaLzR+RTlz4IxC0e+JokyrUPo0zgov4PozCNusn+Imd9k4Gf1d7vXGuxg+qXKtsiXWQiDOnYj6ezXYB4ibFJd6Fcpg1dB0/dMgeQglOFHUsevg9aHwDpzfSiAef3yJBFy2If0xBy7u4oOIc+FKoZDJQJIUgOr6cA4sjG3kXT8FEBNwhmEYFUciBw6coCtHAWcOnFFochlC2Xcg3EStLwPR6I8vm1YCySbBPg3TXcGFznyEUCZoIeBTE6FK35BX5EGk9KtQBh3TvUnCDn1GRlw+XzIx1jTnYGHdk6Txdz7xP8NEgrQzgUuWMwfOd/cSXHyY2wJ7nil+n8SkAi6kAzc8CDo86sBNqHL/9z2WA2cYhlFZJHLg/MflKOAGTMAZBSaXIZQAPXsjHHNq+P3HBFwWk1R/EpwsB07ETZDjnZ5csG/bweGTPlEdON+hqG10E+Mo/fcKSX/gs0rnwPUdgJGh1CGUXbtdU2efooRQBloJxNPRCojrA+eTMwG3y+WWJbroMXcloMUPo/RDKION6sETcCGc82B+p0/DdHPgDMMwKo5EbQTACbhsq34Vg0ELoTQKTF87TKg++CJIVPyJd5gwykwqX+ZEwCUIcYunaW7uQyiTtRDwqakPL+CGekd/H/zbUr1Y5X9WMiF54Q+f7jRuWtMc5850vxzYpgghlKlaCXS0wqRXjC3nP2GCy8vOVsB1esV3/AbhQeYe626LXcgkWwfO/x5XBwTcxGkm4AzDMCqORG0EwJ0ASnVSk4qBHkCyn0wbRlj8cv6JJoZR8JsvhxJwXknxqFUoYayrExU/xC2VgJs8O/chlMlaCPhUT4zW6Nif4Pr90ko1jNIXcLNend6Bi4mxFDlwMNYd7W5zIa/xjk8+SdVKoKM18Xeruj77HLhk+waYOBVmHAM7HsvuGNnSn6yISb2rTqmaevuhBBE1E6dZGwHDMIyKYzBQkS1ITUN5CrjBHjf2bCfThhGWvvbswych4MCFaCWQSdimHzqWlQPX6sI2U10g8UMo0002o5CshYBPFAduMNAnq8abKJdqIRPfaZ2zIn0rgXRumi9eguK6e48TfIX+vUzWSqBz19gWAj7VEXIck9GxK/WFh3ktzoHL5fc2KjEHLr6NQJ27TRdG6YvcgwRcCTpwInKWiDwtIltF5LoEz18jIltEZLOI3CciRwSeGxaRjd7fXbkcvGEYRlkw2ONOjhPifnKjTIhKiYHu0avqhlEI+tqzr0AJo85JGAcu0yqUkH0IZbICJj5Nc51bkMvCCclaCPhEceCGAg5cqfe89D+rOSvStxJIV5AkUYXQVFUr80myVgIdO5M7cNlUTx0Z9sRhktxNcGGU3W3Q/mLmx8mWpG0EvO/rcJr3YCjBBdmJJZgDJyJVwC3A2cAiYI2ILIpbbQPQoqrNwJ3AfwSe61XV5d7fOTkat2EYRvkQbPoZpKahPHPgBrot/80oLP0duXHg6qfAhJqQIZTtbt1grkuY/UN2VSg7diZvIeDjT5JzGUaZrIWAT5SiS0EHzp8ol2oz7752l/926FL3OFUYpe/cNsxI/HzDDPedOUjAFTD/zSdRK4GBHvd6E32/auqzOx91t7n8v3QOHBS3nUDOHLi4HLi+9oIW6gnjwB0HbFXVbao6AKwDzg2uoKr3q6r/n/lnYF5uh2kYhlHG+CGH8ZRrDtxgj1WgNApLrkIoRbxecCFDKKPm3dU2glRl58B1pglDg1GHLpeFTFK1EIDM2ghAoIhJiYZQ9nc4d3fG0e5xqlYC3W3ObamqTvz8hAleeGt8CGUxBFyCVgJ+i4qEIZQhqzAmI1GD8HhesdjlA+4sYh5csiImVb6AS/MdT+jA+c28D2Q/vpCEEXBzgaDXucNbloz3APcEHteLyHoR+bOInJfBGA3DMMqbVA5ctknjxcAcOKPQ9LVHKyaSisaZ4atQRhWNIm6cmQq44UHoejnxBDuI76DkUsClaiEAXhuBkC7amDYC3m9FKTtw9VOce1bXlMaBC+GmTZ4zKpRUixhCmaCVQCqRle0FxTDVU6trYfayIjtwne67PKFq7PJYK4UMHLiG6e62gGGUYQRcoktPCbMPReRioAX4fGDx4araAvwj8GUROejyjohc7om89W1tIX5UDcMwyomkAq6+dPNCUjHYYzlwRmHp64jWjy0VjbPCh1BmIhrrp4zmz0Wl8yVA04dQTjrEhf3lSsClayEA3u9VBAcu1kbAczpK9bcu6LQmK/zhE8ZNa5ozKpQGupyjUwwHLlErgVQiK9scuFTuXpB5LbBro7tYUQwGuhNHkFSHdOD873FCB660BNwO4LDA43nAQb8YInIG8HHgHFWNfQNUtdW73QY8AKyI31ZV16pqi6q2zJpVhC+5YRhGPhnsGZ3EBKmJUBSglBhI8noMIx8MD7rwu1yEUEL0EMqo1GXhwPmT4HRFTKpqoPEVucuBS9dCALwiJmFz4BK0ESjlKpT+5zxjQQgHLo2b1jRntEJoMXrA+SRqJZCqSXx1ljlwHTtdr8Z0r3Xuse68t/vJzI+VDUkFXEgHbihJDhwUtJVAGAH3KHCMiMwXkVrgQmBMNUkRWQF8CyfeXg4snyYidd79mcBJwJZcDd4wDKMsSObA+SFJxSypnAmDVoXSKCB+EYZcVKGE0RDKdP93mRZOqZ+SuYBLNcGOJz7XKhvSVaCEaA7cmDYCZdAHzv+cpx+VupVAWAE31Ov6CKZr/J1v4h3Fzl1QNyVxoZqaLB24Dq8CZXy15Xj8Qiatj2d+rGwY6D64gAlEKGLi94ELnANL0YFT1SHgKuBe4CngDlV9UkRuFBG/quTngUnAf8W1C1gIrBeRTcD9wOdU1QScYRjji2RFTPwJTjYnzWIwkOT1GEY+6M+gH1sqGme5CXY6RyjT1gX1UzKvQhkmj8inae7YhtHZ0Ll7dJ/JqGlwJdZHRtLvL1jEJFaFslQduKCAW+C1Evj7wesND7oJetocuEB+YrrG3/kmvpVAqkbbWefA7Qx34WHK4c6pO1CkVgIDXYkFbMyBS1fEpG/s+lAUAZekjM5YVPVu4O64ZZ8M3D8jyXb/ByzNZoCGYRhlT9IcuEB/pJoIpcqLjVWhNApJJg21UxFr5t2WvGS+f9xiOHDV9aMTwlRMng3PP5TZceLp2eNy6lIdNzjBTeXAjwzD8MDo71tVravMWS4OHDjXaubRY9fr2etuwzhw4MR1MUMoYWwrgfomT8AlEVm5yIE7ZEn69SZMcN/dXBbgicJAN9RNPnh5da27zaSRd/0UQErLgTMMwzCyJJ2AK6c8ONXkOQSGkQ9iAi5XIZTeZNqfkCdieNAJjkwKp2Qj4PwWAmFaFzTNccfJhbPV3eb1MKtKvk7sglMalybeoRBxvxelWIVyZNhVJfSdVr9yY6JWAmHFWEzA7RzdpqFIDlx8K4HOXcmLjFRHaBMRj6rXgD6EcwxOwOWyh2EU+rvS5MCFaCMgVS4P1WdClfu/7ymtHDjDMAwjG5KV3a8OOSEqJYYHXLPWcRZCKSJnicjTIrJVRK5L8PwRInKfiGwWkQdEZF7gucNF5Lci8pSIbBGRI73l3xeR7V7qwUYRWV64V1RG+OGIOXPgvMl0qkqUsWNmGEI50JlZU9+O1vRV/HyCTk+2hCmPH5vgpvm9ijkUgd+ImobS7APXH/fdStVKIKyAm3Sou+3c5XLg6pqKF2ERdBSHh6Brd4oQyrrMBVxfu/t8wwq4pjnFdeCyyoHrS3xBduI0c+AMwzAqirQhlGUk4JI1Qa1gRKQKuAU4G1gErBGRRXGr3Qz8UFWbgRuBmwLP/RD4vKouBI4DXg4898+qutz725i3F1HO5DOEMukxD2R+TN/NyaSVQKoQt3iCTk+2dO9x4iUVsd+rkGXWg6KltqE0Hbj471aqVgJhC5JU17oKob4DV6z8NwgIuG1OvOlI8u+XXxU5k6JasRYCEb67fqXOQjOQpQM32DM2/82nYboJOMMwjIphZNgl/qcqYlJOAi42ORtXDtxxwFZV3aaqA8A64Ny4dRYB93n37/ef94Retar+DkBVu1S1BGeyJUzOBVwIBy7emYmCv03UMErV0RDKMPhOXWeJOXCxEMrARauaxtLMgUvktE4/Ko0DF0KQNc32cuBC9I3LJ8FWAr7jlaxQTXWdE3iZ9GeLNQhP0/7Cp2mOc+wyDTXOhqQOnPf9Hk5SgdRnyBw4wzCMyidWcjhVDlwZCTj/Kvo4cuCAuUCwZNoOb1mQTcBbvfvnA5NFZAbwSuCAiPy3iGwQkc97jp7PZ72wyy/5bXeMOPo7AIHaBIUHMqFmottXql5w/sQy0yqUEL0SZc9eN3kMHUIZqHaYLWEEXKwdQDqHwv/Ni3fgSjCEMtHFgRkLErcS6G5z1RPD5EVOnjMaQllMAQejjqKfc5bMJavOIic7Vj01pAPnjyEXFx+iMDzoLqimDKEM8f1OKuAsB84wDKMySCXgyjEHzs9jGV8OXKKKEvGxP9cCp4rIBuBUYCcwhKv2fLL3/GuAo4B3edt8DHi1t3w68NGEBxe5XETWi8j6trYUrlGl4pfzT9dfKgqNM9KEUGbh+vluTlR3IeaQhBRwtY1ufNkKuKEBN9a0Ai4bB66hRB24BJ9zslYCvsgNW2DGD6FMF5qab/xWAh1pmsSHFTCJ6EgjDuOJhf8WOA9uoMvdJroAWRUyB26oL3EI5UQLoTQMw6gcUoUclmMIZSwHblwJuB3AYYHH84AxMw9VbVXV1aq6Avi4t6zd23aDF345BPwCWOk9v0sd/cD3cKGaB6Gqa1W1RVVbZs0q8tX8YpBpOf9UNM4KWcSkgCGUUQUcjDo92dDj53alCQ0Me8EpoQNXolUo+xM0iQ8W/gjSvTd8PlvTbDeZD+Ns5hu/lUDb35xIaZieeL1sqiJ3trpKm9UhgwiKJuBS5HBXiQnh6QAAIABJREFUVbvqktk4cH3tmRUvygATcIZhGPkkZQhl/dh1ygF/ElYzrkIoHwWOEZH5IlILXAjcFVxBRGaKiH9O/Rhwa2DbaSLiz+JOB7Z428z2bgU4D/hrXl9FueL3sMoljbPChVBmWoUSohcxSRfilgjf6cmGsNUVw/5eDSXok1WqVSiThVDCwa0EooixmMulJSDgvNfzwv86YZnMQfRdpXQhsono2BU+fBKKF0KZrghXmF54SR04v5n3gczHFwETcIZhGPkkpQPnLSunHLjB8efAec7ZVcC9wFPAHar6pIjcKCLneKudBjwtIs8AhwCf9bYdxoVP3iciT+DCMb/tbXO7t+wJYCbwmQK9pPIiLw7czBAhlBnm3dVlEUIpE2DSIeG38YtlZENYARelSh+MDaEs9SqUQQcuWSuBKAIuKMKLWYUSRh3FPc+kzq/MNoQybAET/1gNM3NTQTUKsRDKBDlwEK6VQioHDgoWRlldkKMYhmGMV1LmwJWzAzd+BByAqt4N3B237JOB+3cCdybZ9ndAc4Llp+d4mJVJXztMPSz9elHwHbiRkcS5df2e65dJ3l3GAm6XE29VEaZmTXNdefjhwbGNhaPQHTKEMmzId6wPXMClKOUqlLWTxr7nyVoJRClIEgyDLboDN3/0fiqXLFbEJI0DlYjOVpjXEm2bXFx8iIrvwNUlE3AhmpkP9iZvIwAFE3DmwBmGYeSTmIBL4cCVk4AbHJdVKI1i0t+eWTXIVDTOcg3p+5KEO/W1Q12Grl9VtXPuolah7GyNlv8GntOjTsRlSugQypA5Un5EwUEOXHdx+n6lIpm7G99KYKDbRR+EzoErIQFX2zjqCKb6fsUcuIjno8E+V0E1igMHbv3OEsqBA8+By6KNAJiAMwzDqAhiIZQJfvCralzIVDkJuIFxWYXSKCb5KmICycMosz1mfVNmIZRR8t9gdNKcjZPR3eaKW9SlCRcNXcQkkQPXAGhm4Xn5pO9A4s85vpVA2CbePnWTR8Nviy3gYDQPLlUIZU2GDlxnxBYCPpNnF76ISX8hQigL00rABJxhGEY+SRVCKeImNqU2qUnF+GzkbRSLkRGviEkecuAghYDL8pj1U5K7e8noiNDE2yfWCy6LXCI/NDBdefzqOkBCFDFJ5MB5jkep5cH1dyR2d6cfNbaVQFQBB95nKcmrPhaSGV4eXBgHLuoFRV/ARb74MMc5d5kUTcmUVG0EwBNwYYqYmANnGIZR2aQTPNX1pZkbkoyBbnfyymVPLsNIxkAXoPmpQglpHLgsjlk/JVoVyv4uFyoaOYTSWz+ban7dbeFCA0W8HKEwDpyMLSkfCxcvsUqUSUMoPcfKz4OLhZlGKEjSNNsVRJlQld0Yc8H0MAIuQwcu1v4iaghlDr67UQlVhTKMA5cgB65+CiAm4AzDMCqCVA4ceOW1y8yBs/w3o1Bk01A7FTEBl6SVQLYhlHURQyhjLkZEAdcw3YU/ZhOKFqW6Yk19+t+rIS/ELOjo+VVrS82BS5UDB6N5cL6Aa4gg4I55I7zqrOzGlyuOWgWHLoVZr0q+TqY5cDEBl0EIJRRJwKUKoUwhYIcHXe5sIgduQpX7LlkVSsMwjAognQNXU24OXM+4aiFgFJl8CbiJ0wFJ7sD1Z5sDNwX2PB1+fT8EMqoDJ+JV88tGwO2BWQvDrVs9MYQDl6BKn983suQcuCQ9BhtnOhG+NwsH7sQPZj++XDFnOVzxcOp1Ms2B62h1gihqoaFY/mYB8+AGurwIkiSuaHV96uJDiZrUB5k4DXosB84wDKP8GewFqUpe4rtmYpnlwHWPtybeRjHxwxBzXYWyqtq5V4kEnJ93l80x66dEq0LpFyGJKuDAq+aXoYuhGj6EEsI5cIMJqvSVogOnmtyBi28l0L3H/e5VcvRBpn3gOr3iO+lyKOOJ5W8WWMCl+gyr62A4RRVK/71J1EYA3G+KhVAahmFUAIO9zn1LdnKrnmgOnGEkI18OHIz2gotnoBOXd5eDKpRhy+b75dSjFoLwt8m0iMlAt5uUhhZwIYouDaVy4Erot26g24XDJfucg60EoojcciXWlzSigOvIoP0FuAsktZMKH0KZSsBVpalCmaqqNDgHzgScYRhGBTDYk/zHHtxz5ZYDZxUojUJRDAHnO2fZhlDq8GjOTTo6WqF+amYXR5rmOAcvkx5rYXvA+YQpupTSgSuhEMp07m6wlUCUPMFyparGRYtEdeAyqZ4K7qJmNhcfMmGgO3n+G3hFTFKEkMZaZKQScBZCaRiGUf4k6xnjUzOx/PrAVXIYkVFa5EJMJaNxZuIQyphozDKEEsJXosx0Egxuu+H+zHJvopbHD3PBaSjBb16sCmUJOXDpLg4EWwn4rRYqnTBVGIOMDEPXS5k5x+DlbxbSgeuCulQCLo0Dl6hFRpCJFkJpGIZRGaRzrGpCFAUoJcyBMwqJP8nOdQ4cuIqCKQVcllUog/tKR8fOzAVcrJpfBrlEUYtzhG0jEB9CWYp94NIKuEArgfEQQgkuxzGKgOtug5GhLC4+zC1wDlyaC5ChHbgURUz62p2wzTMm4AzDMPJJOgeuuhwdOBNwRoHoO+AuGFTX5n7fjbPc/ofiihb05yiEEsILuM5dWbgYfjW/DJyMqCGUoYqYJAgbL8U+cOncXb+VwN7noGccOXBRQvpjLQSyuPjQ9VJBBA/g+i2mK2KSlQPnNfOO0kIkQ0zAGYZh5JPB3tQnjLILoeyxKpRG4ejPshpkKnxHpWfv2OW5cP3qp3r7ChFCOTwIXS9Hb4TsE6vml0EuUdT+ZmHaCAwlcOBqJgJSXg6c30qg9XHnMo0XARfFgctWwDXNce9tsnYeuSZMDtzIUHJBGcaBg4K0EjABZxiGkU/SFjGpLy8BN2gOnFFAsm2onYpYM++4yWNsYj81833XRwih7HwJ0OiNkH0mHQIyIbNqft17nEhJNiGNJ0wOXKKoAxHnwpVUDtwBd5vs+yUC0+fD3x9xj03AHUymDeh9fOFXqDDKtG0EPKc/WRhlOgeuYbq7LUAenAk4wzCMfJK2iEmDOylkUkGu0AwNuKuT5sAZhaIoAs4PrctBERNfJKQi20lwVQ00viJzBy5KbleYiIFEDhy4Cz/lVIUSXB5c+9/d/fGQA5cuhDCejlaYUJ25uI3lbxaokEmYHDhI/h6EaeQNJuAMwzDKnnRFP9KdMEoJP3/FHDijUPR1ZCekUhETcHGtBPy8u6qazPfti4IwVSh94ZVpGJq/baY5cFEm32GLmCT6zaspsZ6Xfe2u71cq93HGgtH748GBq5mYuohHPB1eE+8JGcqJWP5mARy44UFXrbV2cvJ1Ys3Mk7wHMQGX5BwYE3AlEkIpImeJyNMislVErkvw/DUiskVENovIfSJyROC5S0XkWe/v0lwO3jAMo+QJ48D565U6fv6KVaE0CkVeHTjPUUkUQpntMWvqnTgIE0LpC69sBVymIZRRhEnNRBgeSF10Yqg3sSiqaSwtBy7M5+wXMoHxIeCq66KdizpbMy++A+49nVBdGAHnf/eyceD85YkcZigtB05EqoBbgLOBRcAaEVkUt9oGoEVVm4E7gf/wtp0O3AAcDxwH3CAi03I3fMMwjBInbRsB70RQDgLOv3pufeCMQpFPAVc/BSbUHCzg+jtyc8z6KSEF3E43IZyYxfQo04bIUUMo001wR4adwEuUI1RbajlwIdzd6QEHrmFGfsdTClRn4MBlc+FhwgTvu1sIAdflbtNVoYQUDlyaRt71UwApDQGHE15bVXWbqg4A64Bzgyuo6v2q6v9X/hmY591/I/A7Vd2nqvuB3wFn5WbohmEYJY5qiCImnrgrhxDKMFcwDSNXqOa3CqWIcwAOCqFsz80x66eEq0LptxAQyfxYTXPcuKM4XCMj0cvj+79lyQqZpMoRqmkorQtVURy4idOhqjr/Yyo21XXh+5KqZteA3mfy7Mx6GEYligM3nKKIiVQlD6+eUOW+UyUi4OYCLwYe7/CWJeM9wD0ZbmsYhlE5DA+AjqTpA+c7cCV0ZToZgxZCWfaour5W5cBQn/sfypcDB859ykcIJTh3J5QD15p5CwGfWDW/CGGUvfvd71MmAi7ZJD8WYpbIgctRCGWueoaF+Zz9VgLjIXwSouXA9Xe4vOhsBVym+ZtRiTlwqdoIhHDgUp3PwTnpJdJGINEloYTl0kTkYqAF+HyUbUXkchFZLyLr29oK1AvCMAwj34QRPLEcuHJw4CyEsux54HPwjdcWZIKRNen6dOWCxlmJq1AWNISyNfMWAj7+JDqKkxFr4h0lhNJ34JIIuLQOXJYXqvo64IuL4NHvZrcfCOfuisDMY2DyodkfrxyIkgPni65scuDAE3Ct+a/E7F88qEsh4Kp8AZcsB643ef6bT8P0knHgdgCHBR7PAw76hRCRM4CPA+eoan+UbVV1raq2qGrLrFnj5CqHYRiVT2wyk6YPHJSJA+edAM2BK18WneMmJ4//sNgjSU+snH++BVyCEMpcVL6sn5K+CqXqaAhlNkzOwIGLCbgoDlyanF1/4pvoN6K2IftG3pt+Al0vwY5Hs9sPhHdaz/smvPmL2R+vHKiuD+/A+RcnJmbRLxHcd3+wO1zF1myIVMQkRRXKMA5ciQi4R4FjRGS+iNQCFwJ3BVcQkRXAt3Di7eXAU/cCZ4rINK94yZneMsMwjMonXclhCIQklZMDZwKubDlkMRx5Mjz6HRgeKvZoUlMQB84LofSv/qvmLoSyLkQIZc9eFyaadQilJwCjFDLxBVxDBg5cuj5ZiVyKmsbRi0CZMDICj3zL3c9FGHDYz3nWK2Hm0dkfrxwI0ybCp7/T3WabL1qoZt79UUIoU3y/y0XAqeoQcBVOeD0F3KGqT4rIjSJyjrfa54FJwH+JyEYRucvbdh/wbzgR+Chwo7fMMAyj8omFUKbKgZs4dt1SJubAjb8QyhDtdI7w2uhsFpEHRGRe4LnDReS3IvKU13LnSG/5fBF5xGuz81PvImn+Of4KaH8Rnr67IIfLmEKFUA71jl6dH+qDkcHChVD6k9ZsQyhrG93xorQS8J3HXDpwqUIos3XgnvsD7HvONS3fty3z/YBzWIb68tdjsFyproeRoXAXdwY8AZdKEIWhUAIuVBXKNA5csib1QSZOK50+cKp6t6q+UlUXqOpnvWWfVFVfqJ2hqoeo6nLv75zAtreq6tHe3/fy8zIMwzBKkFAhlGmqupUS49SBC9lO52bgh147nRuBmwLP/RD4vKouxFV29iNV/h34kqoeA+zHFQHLP686G6YcDn9ZW5DDZUx/gQQcuGqMMCq4clWFcqgvdUhaTMDloL7b5DnRJsE9ewBxOTthSefA+e5NoiImNY1OHA8Phj9ekEe+CZMOgeMud2MPk1+YjFh4bpbhf5VGTZoqjEFiDlyKxthhKJiACxNCGaKRd1oHbrr7buaq2E4SMmydbhiGYaQlVBGTMnLgBsZtDlzadjo4YXefd/9+/3lP6FWr6u8AVLVLVXtERIDTcb1TAX4AnJffl+ExoQqOey88/xC89NeCHDIjcimmkuELuO44AZcrBw5StxLwi45kmwMHo8UgwtLd5nqbTagKv01NuiImfg5cEgcOMqtEuWcrbP0dtLwHZr3KLcvGhSuEu1uOxKoih7igmCsB53/3M2lEH4WBEBEkYRp5h3HgILsLDCEwAWcYhvH/s/fd4Y2U19fntST33r1re+3d9fbeG72TBJYSAqGEEiB8IQkhJIQkvySUBEJII5AAobfd0FlCZ5elbO+9F9u77lWyZVttvj+uXkmWZ0Yzo5Es2XOeZx95pdFoJMsz77nn3nMiBTUKXDzMwDm76eKlZsE3NKAkEmcHgMu8P18CIIMxlgdgHIAOxthbjLFtjLE/exW9PAAd3jEFqX1GDjOvJZVk45NRe0nViEoLpTecmc+D6anM+AiczELOWgewBFKWwkVmicoWymb19viWUAtcOQWOO+5qKFZt+g+Frs++HsjzhmuHMwfXF4XiQDwiFIEJBJ8pC5fAmZNoDjMaLZTmFPk8P7O3iz1cExMg4k6/BoEzYMCAgUhBiQIXTzNwDvtwVN8AZZE4dwE4jTG2DcBpAE4CcAEwAzjF+/hcAKMBXK9wn/TikYjaSc0Fpl0B7Hw9diMFeq20aA+1YAoHPgWOEzhOGnVqoQT8ZEEM1noib3qERGeOBLoalZvTdLeoixAAQp+vfAqcRA4coH4OrtcKbHsFmHIpkFEE5FTS/YYCpz9UETgrXQ/0KOhllkSHwIWKwNFDgeMtyRE2MjEInAEDBgxECkoUOJMFYKb4mIFz2odrBlzISBxBEOoEQbhUEISZoEgdCILQ6X3uNm/7pQvAOwBmAWgBkM0YM0vtM2DfkYnamX8rKSaxGinA7fyZGNfVCdyB0UfgOuhWLxdKQF6Bs9Xp0z4J0H4ED5E4JQhHgZM6X7lkznk+BU5lC+WOZWSYMf9W+n9iKs37GQROf4RSWAPRZwtffePIGKEuw1ALHN2hr18mPWbgvAqcQeAMGDBgIE6hJEaAMbogKA1PHUw4uoerAqckTiefMcavqfcAeDbguTmMMb5SPhPAXkEQBNCs3OXe+78H4N0IvoeBiPVIAb3s/OWQmEouenwGrk/H7DmlLZTcxCFccCMUpUqGFgLnMzEJMQMnplL4ZuBUKHAeD5ntlM4FRs723583JrwWSj2V1qEEtTNwehE4tfObWuDoDu2YmZAAmBLDjxEADAJnwIABA3ELJTEC/HGl2TuDCad92DlQAorjdE4HcIAxdhBAEQDu2OwGtU+uZIztArVO/sf7nLsB3MkYOwyaiXsmSm/Jj/m3UqTAwQ+j/tIh0WeNjkLCs+AA/V0oAXkTE2u9jgSOm0EoWAi7HPReVRO4JABMeoEvd87j5hFqFLgjq4DWw8C8W/vfn1sZpgIXhZD4eESoHLRAOLrCjxDgyBxBmYhKQ8S1QEkLJUAk1u0Qf8zVKz7fGQgfgYtsa7oOTdcGDEigpx149TvA0n/7h44NGBhOcMoM9AfCHC8KnH1YZsABFKcD4IOg+34b8PMb8DtKBj/3UwDTRO4/CnK4HDyM80YKbHgSmPitQT2UAejtjI7JRFpBfwKn19xdcogWyr4umo/Ti8BlqLBj57EJamfgGJMPe3b1AmCkYgRDiwLHowMmBZm+5o7xRwloIWG9nWQeoxcBGSoIFRMRCL0VOIBMeHIq9NlnMBzdyo7XnBRCgQsxA5ecBYAZCpyBOEbTPqB2A3Bi02AfiQED0tj/Pv2LBJx2r2tjiFNtvLRQOhXMEBiIL5jMsRspEI0WSsBL4HiMgFf102PuLjGdSIIUgeOOkRk6EbjUXJrhUULgOGFVq8AB3vNViBYzsc9PrQtlYHSAOYgQ5o6mW60qHC8ORHK+Mh6hRoHrs+lXYOFzoJFso1QyAwfQ35CYEuh2AoI7dEE2wUTnEIPAGYhb2FvpNsJfYgMGwsKqPwDLrwa2L9N/30r65QGq6MUDgXMMzxbKIQ9fpECMBXv3DlILpV5zUYzR8fdJtFBaT9Jtpk4mJowpjxIIm8DJKHBSLn0+F0qFLZSB0QHBCDdKIFrtufEGNbE2uipwKuc3taBPYcunlALnm2kPocAB1EZpEDgDuiKag+rcmtogcAZiGbZ6qpK/+//IUl1POBXa7ltS42cGbpi2UA5p+CIFXoutSIFoK3Aej/6vmZQprcBZvUQrU8f4v8yRChU4jS2UgHwLpbNX+pynRoELjg4Ihi9K4FjofYnuP0rfrXgDV+AUm5joNQMXDQVOxQycmALnkjHoCUZqrpEDZ0BHbF8GPFROF+logCtwsbQgMGAgEK4+GjRe/BNg1GLg7VuAPW/rt3+lCpw5XhS4bkOBG6qItUgBt5NadqNF4AQ3RQjorcwkZ8m0UHoXq3rFCPB9DWYLpUtmRkhNDlxwdMCAffEoAY0KnEHgxKF0Bk4Q9FXgkjKpOKgmiF4tlLZQSipwCk3JgKgocIaJyXCBIABr/kFfwLduBhp3A2f9Tp8ARin0GAqcgRiHrYFucyuBU34GvHI58MZNQIJZH0MHxS2Uqf6KeCxDqaJoIP4QGCmw6EeRvTZwOHuBVff7s9cC4XbSbbQIHECkpreTTDP0QnIWULcNePeHAx87sQVIzta3KJI5Atj3Hl3z5ea7uptp1kfLAlxWgeuRnhFKMNFrhnKhFATx6IBghBMl0Gul876B/lA6A+fqAzxO/QgcY94ogZP67C8Ybifg7gMSlZiYSChwciH1wUjJCS/mQgEMAjdcULMeaN4HfOMvQOMeInNN+4DLno7cBdJoodQfx9cAW54HLn584FC3AfXggbcZJdQKcvXrwEuXAK/fAHznJWD8BeHtX3ELZbLywf7BgttF1sqGicnQxfQriWi0HgEKxkX+9XYuB9Y9Rn9/TKQhKKdCfgGvF9ICwrz1VmbGnkVGG0c+F39cb+fP3EpaqHbUADmjpLfrbqH3rcXEI6SJiUyLWWJqaAXO1kDRARc8LL9dbiWw/wP5baRgKHDiMCsM8nZ00a0SQqQUmSX+tmK9weculSpwYnOaLoWu0gC1MicdUH58GmAQuOGCzc+SRD39KvoCF00GPrwbePps4MplQP5Y/V/TR+CMFkrdcOADYNdrQNk8YN7Ng3008Q+fC1wx3SZlANe8Cby4FHjtOuDKV4Gqc7Tv39mjkMClKBsaH0zwqrmhwA1d5I+n27ajkSdwgkDRBcVTgVu/Glw3wH4KnM4tlEt+Sv+ihRGz6Pbk5hAErlnb/BtA5yveghkMORMTgNrkQhWreGtpdrn8duFECUQroiLeoHQGjhvz6KXAAUR6jn+t3/4C4SOcCgkcHwEKhE+BUzADd8699C+CMGbghgO6W4G971B1lX95534fuO5dqsI9fSZweKX+r2u4UOoP3vK3+iHqPzcQHvjnGTiDkpwFXPsWUDiR3CmlKudKoFSBM6fEvgLnq2AaBG7IwmfNHtnWHwC0UGvaSwHNg23lzgmctT56c3eRQtFkIlAntshv192sbf4NkJ/ZDdU2npga2oWSz/CFysfTGiXgcdN8XTz/niMFX85fKALnXX/oSeAyvA6qHo/0Nke/0GZ0olaBEzUxUaHARQEGgRsO2P4ytT7NvqH//RVLgFs+BzJLafZn3b/0fV1jBk5/2BqAtEKqOq59bLCPJv5ha6B5t5Tc/ven5ADXvgNklQIrw6iiqYoRiHEFjrc9GS6UQxepubSojfDsBgAKaE7JBaZeHvnXCoWUXADMTwTieWFvsgAlM0iBk0N3i3YCFypGQO6cZ0kNXayyKszH0xolwNWjeP49RxJSM2CB6PMqWroqcCMAj0ta3W0+ALy0lKJ/1MKnwCmJEZAgsGoUuCjAIHBDHR4PsPk5oHwhUDRp4OM5FcBNnwDjzgc+vkd7KKYYuALX26lvfIGzF3A59NtfPMFWR8R70sXA2n8CtsbBPqL4hq0BSC8WD9pOzQXGnEmLA0HQtn81JiauHu2vEw04DQVuyIMxUjX0vA6Iob2a2sFnX6/s7yPSMJmpaMOVx3hf2JfOAep3+I1ggiEI4bVQyik0ciYmACkgoWbgrCcBUyKQmie/ndYoAe4Kqlfe31CDnEkNh0+B0ylGAPArrlJGJqvuBwQPULNW/b4NBc5A3OHYaqD9GDDnJultktKBU+6in5v26fO6HjfQ0+G/EErZKGvZ79+nAA8UAn+ZCDxzHvDmzcDK+8nc48jn0c26iyYEgQhHRgk5iLr7gC/+NNhHFd+w1fvn38SQW0nVWq1RGIpbKBUOjg8mfAqcQeCGNHLHRL6FctPTABgwV+a6FG2kFZBxBhD/BG7kbDqXNO4Wf9zRTY+HpcBpNDGxpIZ2oeTnZbHCWiC0Rgn0GgqcLCxKFDhO4HQkwXyUQSxK4MQWclfNKqcCk9riNSdwSginOZnWV8EwFDgDUcWmZ6iKNeki+e24iUnLIX1et6cDgADkeferVxulvY0qh2POBEafTu0iteuBr/8GvPcTktffj+LAeDTRZyVCkFFMrSOzryfS2nJ4sI8sfmFrkCdwvMLbrjEsVo0Cx7ePVThVVDANxC/yxgCdJ0Iv4LTCYaesuYnfpBblWEFaAdBRSz/Hu7lF6Ry6PSkxBxdOBhwgr9C4ekMocApcKK11odsnObQoxj4FziBwolCSSxopExNg4IybIACf/Q5IzQcu/ifdp1aF8ylwSgichALnjK0ipkHghjKsdcCBD4EZV/udhaSQnEXZN7oROK9i4SNwOjlR2r1ZWTOvBi75N3D9/4A7dgG/aQLu2A3Mug7Y9krkW4AGA9xwg7cZnHY3nWhX3Td4xxTv6ApB4HI1tugApBa7epXHCACxTeAMBW54IHc0tSm1V0dm/7teo9y3+T+IzP61Ii0fgLeFOd4X9lllRM6kjEx45qRmBS6V5uo97oGPhVTgFLhQWutCG5hw5I1WPwNnEDh5KJmBUzNTphRpBTSTHkzgjn4OHP8KOPXnwKjF9P2rWa9u31wxVFKANHmDvINHGniHjJzLahRhELihjK0vAYKblBolyKsCWnUicHz+TW8FTurCYzID2WXAGb8mVe7Lv+jzerGEYMv79EIK3N37LnAixMC6gYFw9tL3UlaBq6BbLQocJ2NqFLhYbqHkiy5DgRvayPUaQ0SiCMajA4qm0lx2LCHwmhLvC3vGgJFzpI1MfAqc1hgBiYKTx03hziEVOJkWSkGga51SAhcYJaAUPvUozpXWSEHxDBzT93qQkOB3ouQQBOCze6l1cs4NtL4rnQNUa1XglMzAJVMRyxM0jqMmyDsKMAjcUIXbBWx9gVoNuVNTKORXAS0H9Xl9PjPEX1s3Aue98KRKXHgyiomw7limTTWJZYhZ3i+6nRYen/42+gYYzh7gg19QTEU8okvk8wyGJYUe1/JdUkPgeEUvlqME1FwADcQvIhklwKPlyl/VAAAgAElEQVQD5sdAdEAw+hG4IbCwL51N1/OejoGPhd1C6T2nBReclJzzQrlQ9nbQ44oJnIYoAUOBk4dUC2Eg+mxEgPX+O84o6W9isvddoH47cMav/J1k5YtovpPPMioBv34pcVHmrxP8/Xb1AMxEJDIGYBC4oYpDH9MfwZwblT8nv4qIlh4L8mAFTqsJhNR+5SqHi+8gGf6rIabC8baC9CL/fUkZ1EpZvQY49El0j6d2I7DxSeDIqui+rl7gQ9DpMgocQAsETQqcipZD3wxcHChwRgvl0AaPEoiEAhdL0QHB8F1TGJCo41zPYGGkdw6ubuvAx0IVQkNBSoFzKVAoEtNoO7H2SyAgQkCmsBYILVECnMAZCpw4LCkKcuC69J1/48gc4f8OuF3kPFkwEZh2hX+b8gWkkNVuVL5fRxcVHkzm0Nv6TMWCSKwzRERGlGEQuKGKzc/SCXDcBcqfkz+ObvVoo+QzbzmVAJj+LZTBuV2ByCzxq3CRmuMYDNga6IIT7KI0+3oiGZ/9XvqiGAl0egf+u+I0yiC4JVUKOZWRV+As8aDAGS2UwwI8SkDvLLiOmtiKDggGV6OSM0O7H8YDRs4CwMTn4Lpb6Fqi1U0vlAInNyPEf/dS5zpfiPdIZceiJUqg10qzW0oW88MR5qTQxcQ+q74RAhyZI+g7IAjA9lfIGfas/wMSTP5tSueSElazTvl+Hd3Kr10+BS6IwLl6Ymb+DTAI3NBE2zHg8Epg1vfUnaC4WqZHG6W9lQZBkzKAlGwdXShbKK8n1PtacgfAEoaWCidleW+yAGf9llqTdiyP3vF01NAtb0WMN4i1pIoht4Leo9zchhhUKXASC6JYgrOb/qYDL6QGhiYiESUQi9EBgeAELmmItNUlZ1FRVmwOLpwMOEBagVPaQglIO1HaOIFTqMBpiRLo7TTaJ+VgVqLA2SKjwGWU0LWmqwlY/RCRtfEX9t8mKR0oma6SwHWpIHASsT6GAjcE8fXfgI3/Geyj8GPrC1RFnXWduudll9MCTQ8nSnsbteIwRoRLLxfK7hZlbR+ZI4jAbn9l6KhwPANODJOWAiNmAZ//IXo5eD4C1xSd19MbtnogwULfUzn4ogSOq9u/qhm4EFXpWIDDboR4DxfoHSXgsANbXoi96IBA+BS4IbSwHzmbDK6C56O7m7XPvwHSBSeXAgWOL6KlsuB4+1yo1vZAqI0S6O0YWr9nvWFOGjwCx2cfV95HZP7s34vP2ZUvpO+20nOUo1u5Y6Y5kW4HtFDaDQI3pODoBlb/CVj76GAfCcHVR+6T4y4AshS2IHAkmOjC3apDrpi9jfLnAC+B00uBa1VeOVzyU1Lhvv6rPq892JAjcIwBs66lucdotTR2xHkLZVcjKZqhhrC1Rgn4CJwKBS7WZ+CUDIAbiH/oHSUQq9EBgeDXlaG0sC+dTV0rvNjG0d0SHoGTKjgpCToOpcBZTwJphf5FtBKojRLosxrzb3JQMgPn6NI3QoCDE7jtLwNjzwYqlohvN2ohhW3XbVe2Xz0UOFdv/LVQMsbOZ4wdYIwdZoz9UuTxUxljWxljLsbY5UGPuRlj273/Vuh14DGDgx9T1amjxl85Gkzsf59O2GrMSwKRN1afFsqeNiJuAM2r6elCqZTAZY3058JxshFNdDXp54Tp8Ui3UHJwwqzXZx0KfFFgi1MCF+rz5NAa5u1roVQyAxd7ClxLVx/ufW8P6ju9RNTRbShwwwV6RgnEcnRAIJKzSJEfCg6UHNzIJLiNsrvZf73QAl8LpYQCJ1e08ilwUi2U9crbJznURgkYLZTyUDQDZ4sMCQ50Hz3rt9Lb8XOJ0kBvPWbgnD3xpcAxxkwAHgdwAYBJAK5ijE0K2qwGwPUAXhXZRY8gCDO8/y4K83hjD3vehsC881i1Gwb3WACyWzUlUnyAFuRXUauY2+m7SxAEON0eVbsR7K2wJmTh/Z31cCZlqXKh3FLdhosfX4P/7awb+KDSFkqOJT+l28FQ4T75DfDqFaG3U4KeNsrXkZvX4oQ5GgTO7fJb/carAmcLEeLN4XPl06rAqSBwMTIDZ3e4cNPzm/DcmuP49du7IQgCXQANB8rhAT2jBOq303zuvJtjLzogEIxR0S+9cLCPRD8UTSbFINDIxOMhsqOHAhecFcYX/bImJlyBk2qhrFNuYMKhNkrAIHDyMCeTuiUXTRTJGThTIjD5Uppzk0JaPmUXVyuYg/N4SGBRWpyJEwVOicPFPACHBUE4CgCMseUALgawl28gCMJx72PqVvkxAKfbg5o2O442d6O6tRszyrIxpyLETAx/bo8V2P8xXnediksTvkL15k8xfvJSRc/1eAS8srEGhxttOGtiERaOyYPFpENHa1cT2cyrdNFqsvbi3e11KDyRhos9Lvzy6Xexx1GM1q4+tHY70OfyYFReKiaVZGJiSSbdjsjEiKxkMMbQ53Jj98lObDrejk3H2vDnlgZ80FCG3+zdij+n2XFJQlvIL5vbI+Dfqw/jb58dgiAI+NlrO1CRl4YpI70nWo+HiIya4eusUmot3PoScMrPojt/Ya2jeUJnr3a3Lw4ljoncmVOveUMJCIKArbv3YrbgxgkhH6U9Lfhy30mcMmEEmJIFmiAAXz4CTP8OzV0OFmz1QMUpyrbNqQxDgVNAemJoBs7l9uBHr27DrpOduGBKMT7c3YCP9zTgfKd9WDtQMsbOB/APACYATwuC8FDQ46MAPAugAEAbgGsEQTjhfcwNYJd30xpezGSMPQ/gNABcOrheEASFPUERhJ5RAs0H6HbUovD3FWl893V/IWwowGQBSmb0V+B62qk9NqwZuFAKXIggb0DehbJsvrrjCYwSGDEz9Pa91qGltOqNQAIj9rsUhMgROHMScMNHQMG40NuOWkg5cR6P/Jr36CrAegIYL6PoBR8DALgd/e939sTU+UEJgRsJILD/7AQANX9dyYyxzQBcAB4SBOEdFc/VFb1ONz7e04DdJztxtLkbR1u6UdNmh9vTv8qwdMYI/OrCiSjMlF5017bZ8drzj+FnQh86xl6C43XN6D2yFr9fsQe/+cZEmGXIWJOtFz97bQe+OtQCi4nhhXXVyEqx4NxJRbhgajEWj81Hklmd05sgCGiw9iKpsQYJCTlIdXmQaA5N4lxuD15aX42/fHIQXX0uzLUk4mITkNF1DHm55agqSkd+ehKSzQk41NSFffVWfLjb7zqYlWJBWW4KDjV2oc9F/H1Mfgqy0YXZE8fiuTlzUfPWezD32fDL17fgnm9OQ1bKwBDEhs5e/PS/27HuaCsumj4Cd54zDt/9z3rc+tIWrLh9MfLSk/wXHgkFrrWrDz97fQdKc1Jw+xlVKM7y/v6W3EkE7qu/At+MohLX0w5AoCp20eTw9hXCMbHJ1ouXv27GnQDqG+pREqyR6wCn24MPdtXjP18dRVrdBvw3CWjPnorSzs/xixdWIrekEj84fQwunFIs+/1HRw3w+QNU8T71Lv0PVAmcPVSFVaLAATQHp7TXPvA1AGUKnMlCtsiDPAMnCAJ+t2IPVu5vwv1Lp+CquWX41mNr8PsVe3FObhdMQ0mdUIGATpRzQNfATYyxFYIg7A3Y7BEALwqC8AJj7EwADwK41vtYjyAIMyR2/3NBEN6I1LFrgp5RAq1HaBY5e1T4+4o0lCwa4w2lc8gB1O2k84wvxDsMF0pJBU5JjIC3CCQ2A+fsoQKk0hBvDjVRAoJgKHChEIrAOboBCJGJEQBodlMJyhcBW18EmvfJr7E2PElzlQoFlqGkwImV1GV01QEoFwShjjE2GsAqxtguQRD6XRUYY7cAuAUAysv1r8jXd/bg5fXVWLaxFm3dDiSZE1CZn4aJJRm4cGoxRuenY3RBGkZmp+Dl9dV44ouj+GxfE+44uwrfW1QxQBn738463PPWLvxN+By9KQX4f9deA8/KoxDWPoor1x7A4aYuPPbdmchOHTiEu2p/I37++k50O1x4YOkUXD67FF8ebMZHuxvw0Z4GvL7lBDKSzDhrYiEWjM5DSqIJiaYEJFkSkGgyIdGcgERzAuwOFw41duFAow0HG2w40GiDrdeFjxJrUCMU4rcPf46bllTiynllyEgWT43fWtOO37y9G3vrrTh1XAF+/61JGJ2xBHjoN/j1fAuwZJ7o87r6XDjQYMXeehv21llR22bHNQtGYW5FLuZW5CDPZAf+5MHE0RWYOL4QztOmA58sw2dbD2L1oQ788dIpOHOCP4x65b5G3PX6DvQ6Pfjz5dNw+exSMMbw5LVzcPkTa/HDV7fipZvmw2L3ZsCJXHiarL24+ukNqG6zw+MR8PrmE/jeogr84LQxyM0uA2ZeDWx7CVj8E2rPYIwWFYzB1uvE6gPNONbSjXa7A512J9rtDrTbnejsoZ9LslJw0fQRuGjGCIzMVtgDzVsZmw/4Ti4Olwcf7q7H82uP40hTFyoL0lFVSP/GFqajqjADI3NSYEoI+rPjClzQbEBrVx+e/PIoXlx3HCZ3L+5MBJZ/sQNzy1qwpCqMC3TgS/c68d9NtXhuzXGc7OjB6II0/HpOErALmDr/bOCTz3H/Wfl4cKcbP162DY/kpuKWU0fj8tmlSLaY0Odyo6GzF/Wdvajv7IGnehMuA7B1106scR5CVqoFmckWZKaYkZlsQVaKBaPy0hQVILS/KSLEtsQCbNrfiPLcVIwtlKkm5lQC+96j1lGl0RxqFDjG4DEnY8fROrzWuRNXzx/lV56jiH9/cQSvbKjBD04bg2sX0IL7D5dMwWX/Xov2hA7k51ZE/ZhiBCE7UUAjBt6ebXwOYNCKlbogdwxwYlP4+2k7CmSVqTOlMKAfRs4G1j0GNO4mdcpH4CKgwCkpWvkUOJEWSt91TiWBUxMl4OgGBLdB4OTAf79SDo99NrqNhAKnBuUL6LZ6rTSBaz0CHPoEOO2XfmUtFExSM3CxFSOgZCVyAkBZwP9LAYgMJ4lDEIQ67+1RxthqADMBHAna5ikATwHAnDlz1JBDudfFxmNteGHdcXy8pxEeQcDZE4tw/aIKLBydh4TgBbIXd547HpfOKsW97+3BA+/vw2uba3HvRVOwcEwe7A4X7ntvL5ZvqsXC0kSc1b4DbPoNQEICEsoXAGv+hn+fAdz6Jc1wPX3dHFQV0Re81+nGgx/swwvrqjGxJBOPXjnD99i5k4tx7uRiOFwerDnSgo92NeCTvQ14Z3vojzkz2YwJxZm4eMYIjC/KwJjV3cgorUJlTxr+8ME+PLrqEK5ZMAo3LKrwKYrt3Q48/PF+LNtYi+LMZPzr6lm4YEqxv/0tvUg2SiA9yYzZo3Ixe5REq2nrCbr1Dklb0un2lavH4cefduPG5zfj0pkj8csLJuDfXxzBc2uOY1JJJv753ZkYU+Cv6EwtzcKDl07Fna/twB/e34ffT+vot1+Okx09uPo/69Fk68MLN8zDyOwU/H3lQTz91VG8uqEGNy2pxM3zfoL0bS8D/5g24HAzAFwoMDzmXorXTVchO9WC7FQLclITUZabiqwUM/bUWfGnj/bjTx/tx7yKXFw0YwS+MbUEOWkyixJO4FoOodnWh2Uba/Dy+mo02fowOj8N35w+AsdbuvHlwWa8seWE72lJ5gSMK8rAzPJszCrPwazyHJRZ66mSkk7Et9PuxFNfHcHza46jx+nG0hkj8eMzx0J4IgkjTX24/rmNePDSqfj2nLKBxxUC3X0u7KjtwJbqdmypoZbYbocb8ypzce9Fk3HmhEIkfOmd9xxJlbJzyhnOOus0fLK3Ef/+4gh+885uPPLJAZgYQ2t3/zaEbyTswGWJgL3pGP5SK26Yk2IxYU5FDhaMzsOC0XmYVpoVdotxn8uNPXVWbK/pQOeBL/FTAP9vRR2+8myGOYHhJ2dV4bbTx4irh7mVgMdF4eXclTIUnD2kqpnECygAYO114sNd9Xhr60k85jBhT3Uj3qmpw7KNtThtXAF+dOZYxS3d4eLd7Sfx8EcHcNH0EfjFeeN9988qz8FV88rh2N6FdqcFsdNAElUo6UTZAeAyUJvlJQAyGGN5giC0Qr4T5Q+Msd8CWAngl4Ig6OTdHybyxgB73gJcjvDIV9sRf4ubgejDe47Gic06EjiJNkiXihm44Aw5wG8EFyqbUwxKowT6rHRruFBKg//+xH5HQACBG+TPMKeCvis162jGVgwb/0PmRHNuUL5fn4lJcIHCHncK3CYAVYyxSgAnAVwJ4LtKds4YywFgFwShjzGWD2AxgIe1HqwS9DrdeGfbSTy/9jj2N9iQlWLBTUsqce2CUSjLVTaAX5Gfhmevn4vP9jXh3vf24Kr/rMc3ppVgf70VR1u68cMzxuCnRTvA3ukDJl9CTyojteqMlKNYdstNuPWlLbjkX2vx9+/MQFluKn68bBsONNpw05JK/OL88aItkonmBJwxvhBnjC/EH9xTUN/ZC4fbA4fLgz4X3dLPbiR6F/mFGUl+4uVyAB+3YWRZJZadvgA7ajvw1JdH8eQXR/DMV8dw6ayRmFiSiX+sPITOHiduPqUSPzl7HNKTgr4GeVVAaxhZcNywhOdreWezxme6sOJHi/H4qsP41+ojeGf7SXgE4PpFFfjlBROQbBn4mVw6qxS7T1rx7JpjOI81YiHQ78JzvKUbVz+9AdZeJ17+/nzMKqfl5V+vmIHbThuDv356EP9YeQgvrrPgj5MfQXHPYRxqtOFkezcgCMhKMaOqIBXzbJ/hR7mN+MmN50m+rZpWO1bsOIl3ttfhN+/sxu9X7MFp4wqwaGw+ctMsyE5J9JK/ROQkupHtvbht3boBV366Cg63B6eNK8CfLq/AaVUF/YoInXYnDjd34XCTDYeburCnzoo3t5zAi+vIxvsvKZtwbkIWXvq6Fn1OD55dcwy2Xhe+Oa0Ed5xd5VePUnKwdEwK3mvLw8/f2Ina9h789Owq2dm0zh4nVh9oIsJW3Y599VbwruJxRem4aMZIXDm3DNPLsv1P6qihnB4+w9bViIQEhvOnFOO8yUVYd7QVb2w+gSRLAkqyUlCSlUy32cko238EWAksye/BwdsugLXXCWuPE9ZeF6w9TrR1O7C9tgPrj7bizx/T/ExqoglzKnKxYHQuLptViiKZ9uZgdNgduPvNnfh8fzMcXjOeq9PJQfOiJTNx2/g5WL6xFn/59CA+29+Ev14xvV8hAYB/SL79mDoCZ0kdYNzgcHmw5nAL3tx6Ap/ubUSfy4PR+WlISk7DJZW5uGjpWXhpXTWe+foYLn9iHeZX5uJHZ1Zh8dg8ZTOGGrD2SAvuen0HFozOxZ+/PW1Agevu8yZA2NGHtbU9OM8jDFSIhz6UdKLcBeAxxtj1AL4EXS95KKNUJ8o9ABoAJIKKmHcDuG/Ai0e4S0UUviiB49pbCwUBaD0KTNPJzMmAemSX0zXz5Fb6fzfvZAmDwJkSATCRBa4KF0oxExMrD/FWaWICUJTA/g9Cb8edKg0FThrmEAqcw0vgIhEjoAaMkRtl9To61wRfH/tswLaXqXVS6bgEIP3+Xb0xZeQVksAJguBijN0O4GPQ8PazgiDsYYzdB2CzIAgrGGNzAbwNIAfAtxhj9wqCMBnARABPes1NEkCVx70SL6ULnG4P7v/fXpTlpuLBS6di6YyRSElUN08GAIwxnDOpCKdU5eNfq4/giS+OIDvFgpdvmo/FY/OBZfeRZF/qbTNMzQXyxwO1GzH71Lvw3o8W45YXt+DmlzbDkpCAzBQLXrhxHk4bp+ykaTYlKCacPvDKmlelmV6WjcevnoXjLd14+uujeH3zCSzfVIs5o3LwwCVTMKFYonqSX0WDoVphb6VbH4HzuyMmmU2489zxOHdyMR5bdRiXzy7F2ZOKxPfjxa8unIB99VZ8tPEzLDTB10J5qNGGq5/eAKfbg2U3LxjQclZVlIF/XzMbu0504pFPDuC2TU4AhRhbmI7zTynG+VOKMXlEJi2K37wZqF0vexzleam4/cwq/PCMsdhbb8WK7XVYsaMOK/cPDLIuRDs2es8BKZ1HcNW8Mly3qGIgMfAiK9WC2aNyMHuUX99wewQcaLBha007xn/djSZ7Dh7+iAjNuZOK8NNzxmFiSdDvMCUHiY5OPHv9XPzqrV14dOUhnGiz46HLpvVrSXS4PPjiYDPe3nYCn+1tgsPtQVqiCTPLc3D7mVWYPSoHM8qyRecVAQAd1f6FAdAvSoAxhkVj8rFojEQL51bvtp0nkGhiyE9PQn56/9aGpTPpAt7a1YcNx9qw/mgr1h1pxcMfHcB/vjyKv35nBs4YH3oe62CjDTe/uBl1HT24bmEF5lbkYEZZDor31QAfAd8+fR6QmotFY/JxzqQi/Oad3fjGo1/hngsm4toFo/xEJnDGQqmY4A397O5zYatXydx4vA3bazvQ6/QgO9WCK+aU4dJZIzGjLBvs8UwgwQkkW/DDM8bihsUVWLaxFk99eQTXPLMB08uyccdZVThjgro5NI9HwKGmLiRbEpCTloiMJHM/InigwYZbX9qCirw0PHntHNHiUlaqBe4EB2q6EvDqhmpcu7BC1TEMAYTsRPF2m1wKAIyxdACXCYLQGfDYgE4UQRB49kwfY+w5EAkcgEh0qYREYJSAVgJnbwP6Ov0FEAPRB2MUJ8CNTOwtAJj/+qx1n5aUgQqNq5dGE2S6DmBK9M77iszA2TiB06jA8SgBOXJmELjQ8BGYUArcILdQAmSOtOctKirnBM3Zbl9GZFNt/qSkAtcTviGdjlA0zCEIwgcAPgi677cBP28CXdCCn7cWwNQwj1EVMpIt+OAnp6A8N1WXanWyxYQ7zxmHa+aXIznRhMxkCzkYHf4UmPv9/s43ZfNoTsbjQUlWCl7/wUL8fsUe2HpduPfiyQMWqbqjy2t0kd6fEFXkp+GBpVNxx9njcLS5G3NG5Ui2kAIgAtfTBnS3AmkasmK4CyJ3RUwdaG8/ZWQWnrhW2aCq2ZSAx747E2///TnABTS709FU14lrn9kIUwLDf29diHFF0ieSqaVZeOHGedhT14kkc4L4rFNWKbCnDvC4KdBcBowxTB6RhckjsvDLCyZ45+Sc/ebn0LQPWA/Ykwowwd2Ie781SbUzqCmBYdKITEwakQnssAPFY7H1knPQ3eeSJvepuYC9HRZTAh6+fBrKc1Pxl08Por6zF09cMxtHW7rw9raTeG9HHdrtTuSlJeK788uxdOZITB2ZpVxZ6aylRYE5kX7PaqIEeJXV1UPV4HTpokZeehIunFqCC6fSBf1wkw23v7oNNzy3CbeeNhp3nTtesrXy072NuGP5NqQkmrH8lgX9W35t9bSQCHCU+tb0EZhXmYu739yJ363Yg0/3NuLhy6dhRHaK19o4SdSJss/lRgefm+x2osM7PznleD0KexgW3/sJ3B4BCQyYNCIT3503CovG5OHUcQX95/wsyf0WRKmJZty0pBLXLCjHG1tO4N+rj+CG5zfh8tmluPeiyUgLVs5F0GTtxc9eJ8Mk38uYGHJSE5GbRv8ONXUhxWLC8zfOkybsHjdMHgfyc7Nx70cHcN6UYhRmxM6FLAoI2Yni7TJpEwTBA1LWnvXeL9mJwhgrEQShntHFaimA3dF6QyGhR5QAf67RQjm4KJ0NHPwQ6OnwZ8CFuM6FhDlZfIFrTpGPi2CMVDgxExNrHbXlaSEGgQUHOSfKXm8LpUHgpCGVg8YRSwSOz8HVrOtP4DweYONT1EJcOkfdPsVMTNxOmp00x9cMXNxhVJ7+Vtf9HCkPfEj2orx9kqN8ARlltBwECqkl8KHLBs5cRQxdXiUoQ1zRElM6RJFXRbeth7QROJ8C532uT4HTbm+fl56ES8Ylo3NPGm58aTuqW7uRnmTGKzcvQGW+st/35BEyJ+ysUppxsjVQFpBCMMaQnZqI7NREVCLgOI5XA+uB1DGLgb3vkIVtOJb5tgageKpv0S2JlByfExdjDD86qwqluSn4xRs7Me+Pn6HP5UGSOQHnTCrCpbNG4pSqAvWzZR430HmCcloAak1QQ+Bs9aCONAHorJElcMEYW5iBd364GPf9by+e/OIoNh9vx6NXzexnLCMIAh5bdRh/+fQgpo7MwlPXzUZJVtBJ19ZALaBBC42izGQ8d/1cvLqxBg/8bx/O+/uXuHTmSPQ6PbjDVIyT27bij4fXwNbrQlefC509TtgdbtFjfSKxDdmJybjttDGYW5mLWeXZkoZCAKg1Q6TimWQ24er5o3DFnDL8c9VhPLbqELZUt+OfV82UNTr5bG8jfvHmTtgdLtxzwQTkpyeh3e5Aa7cD7d0OtHn/leem4r6LJ8ub83jbnc6YUoFffeHBA//bh0evUmDXPUSgpBMFwOkAHmSMCaAWyh96ny7XifIKY6wA9AexHYDKUnEEoUeUAH+uocANLnigd91WInDhtE9yWFLETUyUKBSWVHETE2udtvk3wF8kCEngDAUuJLhRh+QMXBfdxgKBK5wEJGWRkcn0K/33H11Fa9hLnlK/T5MF1CIcML/vaw+OncLlkCRwEceet4HMUv9JkaPMWwmo3QAUToj+cdnEFTjVyPcSuJaD/uqGGtjbgASz/487KYvaKsIMmM5jVnSl52PXyU6U56bi1ZvnozRHp37kLG93VOcJVQROEvy9li8gAtd8UDuBc7uA7iZlF7aU7AGf8yUzS1GSlYLlG2uwcEweLphaQkqyVtjqiexmez+z9EJ/8UAJrHVA4UQK9+2o8Q/ZK0SyxYQ/XjIVC0bn4Vdv7cI3Hv0Kj1w+HWdPKoLd4cLPX9+J93fV4+IZI/Cny6aJzlbKhXgzxnD1/FFYPCYfv3xrJznDJpux1F2IEvdJpCWZUZSZjPQkMzJTLMjhc4+pif6f0ywoWvEsEnpcuCvAEEQW5mTA0SX5sMWUgDvPGYdFY/Lw0/9uxyX/WoO7z5+AGxdX9lPUexxu/OGDvXh5fQ0mlmTin1fNkHfYVAJvu1NeTg5uO30M/rHyEDMzCvUAACAASURBVC6fXYpTFbaEDwUo6ER5A8CAOAC5ThRBEM7U+TD1A2OkaoQTJRBPEQJDGSNn0e2JLdT1EE6EAIc5eWDBydWrTKFITJVW4NQ6UHLwNvfWEAWHXq8ZmkHgpBFPClyCCSifD9QEjcBseEpddEAgGKPPIFCBU2LQE2UYBE4tejqAIyuBebcMbInLG0OqU+0GYPb3on9sfBGdFmZWU3Y5tYvJOFHKwt5KnwNXNxISgOSBxEL9fluQnluMVy+dj/FFGZQNpxd4wHdnLdTFHEqAq408kLTlIFB1trZ9dTeTmYCSIdyUHNHPmbs56oIOrxkfJ6TpRdS+oASCQARwxne9BK429HMkcNH0EZg6Mgu3v7oV339xM65fVIENx9qwv8GKey6YgFtOHS3dRm1rAArkiVVFfhqW37LQf8eHnwFbX8RLN86TbxHiUDvwbEn1GwzIYMHoPHzw41Nw95s78cD7+/D14RY88u3pyE9Pwt46K368fBsON3Xh5lMqcdd54oZJqsENBxLTcNvpY7BiRx3+793d+PiOU8UJsoGhgdzR4UUJGBECsYHkLCB/HM3BdTcDxTp0BllSByo0ihW4NIkZuPqQ52VJKI0SMFwoQ0Mq54/D9xnGAIEDqFB+6BP/2E/rEeDQx+qiA4JhTupPYNXkukYJEQxaGqKQap8EaFFXJlIJiBa6GmgeKdyLZYKJyGjrYW3P72nzz79xpOT43Sm1orsVSCVzDF3JGxBA4E7Ib6cUnETljaX33iJul68IfLBbkQKXQyddqdYHPdBBDo6+qjpX4AQF3gr2Vvr7KZhIymyndgIHAJX5aXjztkW4dsEoPL/2OE602/Hs9XNx62lj5GdgbQ3qW3VyK6ntR6na6LT7M4+UwJIsvqgRQU5aIp68djbuXzoF64604vy/f4U/frAPSx9fA2uPEy/dNA+//sYkfcgb0C/TLtliwv0XT0F1qx3/3RTe789AjCNvDP2NuhyhtxWDESEQOxg5h6IEuvRqoUwWNzFRssBNTB3oQulxazsvB0JJlEBvJxWoY6gVLuagRIEzJWonR3qjfBHd8kLyxv9QF5ia6IBgBM94KgmpjzIMAqcWe96miqJU21fZfLpoKaik646uJnVWqXLIG6uddNjbBmS1SSlDqtDdrE/rhxiSM6lKqSeB422k+ePCJHDe1lilChx//UiBEzhOetOL6UTHq3JysAa4jGWXhaXAcSRbTLh/6RQsu3kBPvjxKaHdKR12csaTmBWVBG/RETEyEYWzR121zpIy0BRABowxXLtgFFbcvgS5aRY89eVRnDquAB/dcSpOqdK5tZG3O3kJ6ZKqfDx/w1xcs8BojRvSCIwSUAseIWDMv8UGSmeTS2Nfpz4ETs7EJBQsqQOLVV1NZBKhtYUSoCiBUC2/oVwqDfivW1LXI0fX4EcIBGLkLCLlNeuIXG5/hUSWcNbDwQocVyNjKEbAIHBq0NMOHFlFPbVS1X3eMle7IXrHxWFrIDVED+RX0UXb7VT/XHub33mSIzU3PFLh8ZB6EykCBxAx15PApeTS9yRsAqci3JQrn5EkcJ011KbLT/J85tKmwMjE915GeD9v/RSchWPylEVvcLdWLQoc4DOJCQmnXd3J3pyiWIELxPjiDKy4fQne+MFC/Oe62fImN1rBDQcsfqOe08cXDsc8uOGFQGc/tfBFCBgKXEwgcGZfj+uoVIyAEmVLzIXSFyEQBoELjBKQQq+VCrYGpOEL8pYgcH222GmfBIhsjZxFBG7Hciomq40OGLDPYAXO+3MMKbcGgVOD/R8AHqd4+yTHiJkkLQ9GG2VXE6kheiB/HBlVaKm88hm4QKTkhOVCid4Oqs6lRpLAlepH4OxtfjUsfxyph1pbSG0NZASghJxHS4ELNGThx6XEidJ6km59ClyN/scXCmoUzUBklwNgkVXgpC6YIZBsoaDzSAV9BytwBoYJwokSMCIEYgtFk/0Lc91MTIIVOLsKBS6ohdKqB4FTUHAwFLjQELPRD0SfLfZmCMsXAvU7gHWPa4sOCIY5icY9OLgCF0MxAgaBU4M9b9MibsQs6W0syUDJDKB2Y/SOC6B2lS4dFbi8ACdKtcchNQPX06H9eHg0QUQVuFL9FKGe9v4EDtBuCmOrJ5VLSW5P1AhcQKYxV+AUEbh6LxktIgWuz6rue/HRPcCRz9UdbzC4Cqi22GFOou+IYgWuR6WJSQpdJJTMEkYbjoEKnIFhgHCiBIwIgdiCyUJrE0DHGIFgExOlCpyIC6U1oDtDK/IMAqcLpIKsOWJNgQMo0NvjogLrvFvD358pyVDghgzsbcDRz0l9C1XlLpsH1G2THgCNBHo7qFqg1wxc/li6VUs6+qz0RzRAgculx7S0ZAL+mcLg/eqJrFL6HLlFbjjo6QggcBrJMIe1XvnvNdIEzuMhlTJQgeOzZErMPWx11H5psvj3oZQ093YC6/8F7H1X3TEPOAYv0dTyt5JbqUyBEwRvC6VKBQ5QNQcXNfBquaHADS+EEyVgRAjEHrgqESkC51I6AyfiQmk9Sd1L4VzjfVECMt/XPmvsqUexhgQTkGAJQeBiaAYOAErnAmDaowOCYU4Wn4EzFLg4xP73iZjItU9ylC8A3H1A3fbIHxcHX5SGmwHHkZxF+2pVSeB4m2CqiAIHaFfh7F4CF+kZOADoPBn+vnra/Z9BTgVdmLQSODXOXJEmcF2NVCgIJHDJ2fT++GyZHKz11D4J+FU8pUYmzd7Pr7tZ+fGKwVZP1bWUnNDbBiOnUpkC53ZQy68aAscvDJF0ENUKh9+F0sAwgxJnPzEYEQKxh6nfBiZ803+tCwdmEdMlp0oXysBuA5u3UBkcz6QGialA0RRg33vSnQyGAqcMci39sajApWRTvNfZv9PHHTM4B85Q4OIYe9+hhThvQZDDYBiZdOlM4ABqo1SrwPkInMgMHKCdWHAFTo/KoRT0jBIIbKFMMHldPcNooVSqFiWmEZkKN7JBCsERAgBV6dOLFCpw9f4WmSyVClzLAbpVExouegzeEG8t82K5ld4h+RCOm04NhMcSwwSOv59Eo4Vy2EFrlIARIRB7GDEDuPIVfUh1ODECllQAQv8FsrUuvPZJjrk3AQ07pddfBoFThmACE4hYJHAAcOHDwMxr9NlXsALHr4GGAheHqN8BVJyibNGXXkiV+ngncPkaCBw3KgmegUvVicBFuoUSCH8OztVHLWcp2f778qv8BETtvnralCtwjOkT2SAFX4RAUAU3vUi5iQkfUk/Lp5OhUiOTZu/n1x0mgetq0N5qrDRKQEvoZywTOEc3tdSYLIN9JAaiDR4l0FGt/DlGhMDQhzmFTN08bv99zh5lOVm8EBQ4B2etC8/AhGPad4igbXhi4GOuPiIlhgtlaJhTpMeAYi1GIBIIJrD8ZyPIO87gsFPbVk6F8ueULyAnymgZEvDFs9psKznkVxF56G5V/hxuNiLZQqlRGbK3UN96JIMj04sBZgpfgePkKbBFL388OXqqnYu0abC8jySB6+QKnAiBCxUj4LBT9ZO3UDJGpFkpgeMtqF3htlCGQeCURgk4NWTG+GbgYpDAOe2G+jZcwZ391MzBGRECQx+8lYyf69wuInSKFTj4Z2sFgboz9CBwiWnArOuAvSsGjkPwzonk7IHPM9Af5iTxa5HbRdeDoT5HaE7u33XgNAhcfEKsbSwUyuYT6dAyO6AFtgaqmOj5R8WdKNXMwUnOwIWZT9bdEln1DQBMZrqAhKvA+QhcwGeQP46q2Gq/D7FG4DpqKMoheDGfXhhagbOJuIxlq8iC4wqcs9vviqgFamYKg6FYgeMtlENoBs4gcMMTWqIEDAfKoQ9+vuLKhM/kQaELJeBX4Ho76JypB4EDgLnfp+vt5mf738/z4YwWytCwJIsXnB1ek7dYbKHUEwMUuB4q8MdQF4pB4JSAE7gclQQOiF4bZVcTLaL1zIHyuSeqIXCt9CVPCjpBhjsDZ2+JrIEJhx5ZcKIKnPezbFbZRsnDTdUoRuFGNsghOEKAI72IfvdyLqO+nJ8A8pRdrszExNlLLVyZ3jZXrUYmjm5yIdPaapycSYUExQrcEGmhdHYbBibDFVqiBIwMuKEP3/nKS8LUKBQ8joQrcL4IAY2FtWDkVADjLwS2PNffiIMTuKGuHukBs8iMIwD0ddHtsCBwgTNwCuc7owiDwCkB7/1Xo8AVTKCLXrQCvcOZ65FCdjm59alxT+zxBlgHO0klZZKltFZzje7WyIZ4c+iRBcffoxiBUztTGHMKXG1/B0qOjCIAgn9WUQycwAUqcFllRM6DM4GC0XqYKqqVp9D/tbZRavk8g5GjIEpAk4lJUEtSLMFhNyIEhiu0RAkYEQJDHz4CF6TAKXWhBPznfV6ozByp3/HNv5WKirvf9N/XZyhwihFs4sHBY5ZiLUZAb4gpcErU5SjCIHBK0FFNvzg1IdkJCUDpvOgrcHoiwUQtMK2HlT/H3ire6piQEB6xsLcAaRFuoQSIUFjr+g9mq4WYApeYRvtWGyVgqyfziOCWVDmk5GifNZSDIBC5FSNwvjBvmSgBm4QCB4RWPbkBTMUSutVqZOIjcGEUO3Irgbbj8ttoUuC8ixqpGbiOWmDtY+rnaq11wJvf91dOtcBpN0K8hzPURgkYEQJDH3wxy89XnMgpWeT6FDgvgRPrzggXlacCBRPJzISfM40WSuUwJ4tfi/qGSwtlMsWB8e+OocDFKdqraaGptj2xbD7QvD9yakggbA1kwqE31DpR2tukyYZWAid4lZ1IRghwZJVS3p8SR0Up8PcY/DlocaLk81pqvnspOXRhlMpw0YquJqpIiVXVfQROhlhZ60mJDTzxczfLUEYmLYcAMKB8YejXkYNNh1adnErAekLeVl2LAmcOocCtegD45Nfq3AAB4OBHwK7XgZp16p4XCEe3ocANZ6iNEjAiBIY+fB0D4ShwQS2Ueq5hGAPm39I/UsBH4IwWypAIbiHk8BG4+PoMNx5rw4EGm/IncMM8/hk47YYCF5foqBZXHUKhnM/BbdL3eILh7KUhYD0jBDjyq6hdTG62KRD2toERAhxalaHeTnK3ikoLJQ/zDmMOrqcdSDAPtNnNH09ExONRvi9bvfqqJFf+enWeg+OtpWIhsD4CJ0N8bXUDiROfp+sMQeCaD9AMKn9trTNweri15lZ6bdVljjkcBU6MwNnbgD1v088tKhTxwO0bd6t7XiCcdmMGbjhDTZSAESEwPGAOcs1VpcDxcx1X4E4CaYX6K7bBkQI+F0pDgQsJS4r4tYibmMRRjECPw42bXtiEX729S/mT+PfY7SVwSjMOowiDwClBR422Xv6Rs8nQo1bDHFxXE9Cg8MvG28n0jBDgyB9HilT7cWXb98gpcLnaFDgeTRAtExMgvDk4PgcYrJrlV9EFi7cSKoEWy/twDWOk4JsFFWuh9LbvykUJWEXIaEYJkd1QRiYtB4kAmxPJAjocBc6cHJ6NNF+Yys3BaYoRkFHgdv7XfyFR4wobuH1DGATOcKEc3lATJWBECAwPBCtwapx3g3PgtBQqlSA4UqC3k2Yz44h8DBpCKnDx00L54e562Hpd2FbTjrZuhV0EAxS4HoPAxR16rbQQ1qLAJaYBxVOBGg1zcB//GnhxqbJ5F75ojoQCx6MElMxuCYJ3Bk5GgbNrIBW+EO9oErgwFbjA+TeO/HF0q8aJ0lqvvt0vYgROIgMOoJNdcra8AmetGziknmCi++QIs8dNymWB9/NLLwxvBi6jODy31hwFWXBaYgSkZuAEAdj8HDBiFlWO1Rrh8O0b96h7XiAMF8rhDV+UgII5OCNCYHjAHORC6dKiwPEWSpFrg14IjBTo7fQaquno1j1UYU7pb+LBEYcEbvmmWqQnmeERgC8OKlw7mDiB6/XfGi2UcQYtEQKBGLUIOLlZ+ewAQAu241+TcYecqx9HVwQJXP5YulWyaHR0A26HdF6b1hk4u/cziIaJSXImRSCETeBESCwncEoX4H02alfQqsBpdfyUQkct7VvqxJ1eJE3gPG56TIyMhooS6Kgm9Sl/PP0/rTA8F8pw5yzSC2kIX24xy1U0swoCZ7KQYh+swNWsp9nJOTdQQUWNAufqo8/PlEhFGLVB8hzGDNzwhi9KQIECZ0QIDA/w4hRf4KppG+cLYUeAiYleEQLBCIwU6G4y2ieVItiFkSPOYgSONHdh47E23Hb6GOSnJ2HVfoVrB59Jj6HAxS+0RAgEonwh/RHUbVPxmjX+Njslyhd3/osEgUvOov0qWTTyVkepGbjUXCIkSufpOKKpwAHhZ8FJKXDphV4FRaETJVdW1V7YUsMMTZdCR428Ep1RJN3a2NUECG7xNpmsMvl5smbv51XgJXDpBeErcOGAMVoUyLZQegeeg+M0QsGSOtB8ZstzVDWecpnXCEfFDFzbMao+jz2bPn+1OYQAkW9Xr+FCOZyhJkrAiBAYHgjOreSLfSWL3IQE77nOa7bV06ZfiLcYeKTAgQ8NAqcU5mT6nQZ3gfVZ6XeXYBqc41KJ1zbVwpTA8O05pThjfAG+ONAEl1uBD4HZUODiH+06EDgAqFmr/DmBbnGKCFwTABY5l8Y8hU6U3KBEToED1BMLnwIXTQIXxgycXYLAMUYqnGICp9ExMZItlHIELr1IOkbAJpIBx5FdRu9VSqXmzp1cwQxXgdOj0ptbGaKFUmO1zpLsb0kCvOYl7wDTrqCW7Lyx9FkqjQTghZdJS+lWSxslPx5DgRveUBolYEQIDA/4FIogBU5p14EllZR9X7xMBAkcjxRwOwwCpxSWZCr+BRfc+2xxo745XB68ufUEzppQiMKMZJw1sRDWXhe2VCtYGw1Q4AwTk/hDRw0NvKrJ4QpEegEtPKtVWHjXrKM2PnOKMuJkayByYzJrO8ZQyB+r7Di4Aic3AweoJxbdLfQ7iNYfT6QUOEAlgdMYOp2YTsYgehI4ngGXFYrANYnPbVplLtJZZQAEciITQ/NB2neK13gkrYBMEtTGJGhtSRVDTgUZ+4g5itbvIMdILTMdlqC5gx3LqH109vX0fx4IrzSbkf/djjuXLkhanCgdGiIRDAw9KI0SMCIEhgckFTiFKkWiV4Gz6hDtEgo8UgAwCJxSBBN0jjgicKv2N6Kly4Er59Hc/pKqAlhMDKsOKOjgCTYxMYK84xAdGjPgAlG+kOZYlIZDV6+jCIL8scoVuEhkwHHkjyN1rbtVfjtuUKK3AtfdIr3PSCCrlI5RS/Cxq48Gs1NlCFxXI9CjwOLfp1qp/N0yFl5ouhjsrXSxDaXAOe2AQ+Rz4xdpMQLnC/OWUD1bDvjVN4CKIoD6KAFfS6oOfyu5lUSsuErKcewr4Plv0gD05c+p3685xa94CQKw5Xlg5BwyQwL8pkJKCVzrES/5zQEKJ2pU4LxGA8PchZIxdj5j7ABj7DBj7Jcij49ijK1kjO1kjK1mjJUGPOZmjG33/lsRcH8lY2wDY+wQY+y/jLHYla3yqqgi37BTehsjQmD4wJQIgAUocN7zlmIFLo0UOF9xL0ImJhzTvkPnwUiMmsQBXG4Pdp/sVP4EKQLn6IobF8/lm2pRnJmMU6tozZCeZMa8ylys2qeEwIkozDFWxDQIXChojRAIxKhFpBg07Q29bXcrLVjLF3hbFxXOwEUiQoCDm0c075ffLtQMnFZzDXtL9NongfCy4Dgxk1PgAOXKqiVNW7VLbwInFyHAwS+MYlECtjogwSI+x8hdLcWMTASBFDg+/wZQCyWgfg6Ot3fqosB5nSgD5+D2vQe8fBlVkm/6xO+aqQaWFL+yWL2W/v7n3OB/PHc0AKbcCKf1kJ/0FU0OT4EbxgSOMWYC8DiACwBMAnAVY2xS0GaPAHhREIRpAO4D8GDAYz2CIMzw/rso4P4/AfibIAhVANoB3BSxNxEuxp1L56NNz0hvY0QIDB8w1j8rzNlLs48mi7LncwXO10IZQQUOoPPXLauBM38T2deJUTzw/j58859fY1+9VdkT4lyBq+vowRcHm/HtOaUwm/xU58wJRTjU1IXaNrvMs+FvAe9nYhKHCpyCyuOpjLGtjDEXY+zyoMe+560uHmKMfU+vA48KBIFm4LRECARi1CK6rVYwB8cz48oX0WK/o0Y8FyoQXU2RrSoVT6HbULl0PW0AmL/VLRhazTW6W6JnYAKEFyXA5wBDEjgFxNxWr93yXmtouhQ4uZIlcF5iJeZEafW+FzFTj8xSAEzcyKSrkRaE+QEEzvc6ahU4jS2pYsgNihLY8gLw2nVAyTTgxo+ALI3VZEuAArflOWqlnnxpwOPJ9DtQ6kTZcsjvJFs0hVRLtRl6vkiE4UvgAMwDcFgQhKOCIDgALAdwcdA2kwCs9P78ucjj/cAYYwDOBPCG964XACzV7Yj1RnIWMOMqYPcb0n97RoTA8EIggXP1kvqm9HplSaXikLWOTJqiQQpyKrSPw8QxPj/QhOfXHgcArNqv8PzvcxkNci7us9HvK8bx+mZav10xp3/s0ZkTaP0Q8nMIJLBuJ5mAqXGVjgJCEjiFlccaANcDeDXoubkAfgdgPugC+DvGmMTKNgbR004zM1ojBDiyy2mRqoTAVa+l1oQRM73zLoK885fHQ4vcSBK4jGJSPeRaZwBS4FKypd2JNJuYtEZZgQsjzJu/NykCl1NBSpQiAheG4YbW0HQpyGXAcfDvoBiBs8nYRJsT6Tsm9nlz18RANYub9ahV4HymMDoocFllZPnffgz48hHgvR8DY84Erns3vAUCn4GztwF73wWmf2egeUi+QlMhexuR+EAFDlCvwjl4C2VstY9EGSMBBH5BT3jvC8QOAJd5f74EQAZjjPd+JzPGNjPG1jPGOEnLA9AhCIJLZp+xhXm3kBHE1ufFH49ShIDbI+ChD/fjYKMtoq9jIAQCs8LUGjclplF7diQjBAyg2daHn7++AxOKMzChOAOrlcx/Af4ZsGABIQ4UOLdHwGuba7FkbD7Kcvtftyrz01CZn6aAwAXMwPkiMuJPgQtZeRQE4bggCDsBBE/0nwfgU0EQ2gRBaAfwKYDzdTju6CDcCIFAjFpE5iShgrlr1gMjZ9MXhas1ctX2nnbA44p8X3fxVAUErk1+Vi0pkxa9apQhQYj+DFxGCR2nJgWOEziJRbzJTIsbRS2U9drbSlJylM3ZAXTMqx7wL9TF0FFDFXi5AXBOjMQUHmudvMuYVJQAJ7qiCpxaAtdACw49qocmC5HZjU8Dq+4Hpl4BXLU8/DZDs7eivf1VWihz85JA5FVRUSfUuYR/x7jxSSEncCrn4JyGiQkAMVkh+BdwF4DTGGPbAJwG4CQATs7KBUGYA+C7AP7OGBujcJ/04ozd4iWAm5ubNTqw6oGC8VSo2PSMeBxM29GoRAh8vKcBT3xxBH//TKEhlIHIwJLcX4FTQ+C4AhfOdc6ALARBwM/f2AFrrwv/uHImzp5YhK01Hei0K4hyMsspcLE9A7fmcAtOdvTgO3PFC85nTijEuqOtsDtcoo8D6K/AqQmpjyKUEDgllcdIPHfw0a5g7kcpRi0kZULOhtnRDdRv90cP5CkI0fbN9USYwJVMA5r2yzuQ2VuliQugzVyjz0ZmEZGKSBCDyUxkIywCJyM051f5rfGlIAjhZZap+ZwPfgx8+Wdg7T+lt+msDf13kJxN6mJwlIAgUAulHIHLLpNW4JIy+38OlhQgMUODiUmD9pZUMeSOpvbO+bcBlzypfPZDDryFcsvzQOk8v2oWiPyx/sq1HHjhh59H0vKoOKGWwBkzcABduwJXA6UA+v0CBEGoEwThUkEQZgL4tfe+Tv6Y9/YogNUAZgJoAZDNGDNL7TNg308JgjBHEIQ5BQVRPBeKYd6ttOjet2LgY61HIh4hIAgCnviClL5P9jSi2aYxnN5A+AhW4NQscH0ulHWRNzAZpnh+7XGsPtCMX184EeOLM3D6+AK4PQK+Oqzg2ulToAIUOEEIqcC53B58fqAJt7+6FVN//zF+vGwbTrSHmDfTGcs31SAn1YJzJomvi8+cUAiHy4M1h2WM+fh32e1QF1IfRSghcIqrhFqfGzPVxWD42sZ0IHDlCubgTm4hNY0TuMRUsm2Xa7fj7WrRUOA8Tnkjk54QChygnsBFOwOOQ2uUgD3EDBxAymrbMXky3NNOF0bNLZQ55BYVyvIb8BOnNY9Kq1odNfIRAgDNt6UXDtxHn5UIh9x7yS4HOk8OdGrlDpTBpCu9QJsCp2erzil3ARc9Bpz/oPrAbilYkum70Xqov3lJIHxOlCFU3JZDRKgD1ZCiKepbKLkL5fBW4DYBqPK6RiYCuBJAPwbDGMtnjPEvwj0AnvXen8MYS+LbAFgMYK8gCAJoVo7PjX8PwLsRfyfhoupcMvHZ8NTAx6IQIbD2SCt2nujE95dUwuUR8MaWMCJfDISHAQqcCgJnSSOnZ73PywYAAPsbrHjww/04c0IhrltI14AZZdnISrFg9QElBC4oB43/7HGKEriDjTY8+ME+LHxoFW54bhO+PtyCU6sK8PGeBpz5ly/w8Ef7YetVoPyFiZauPny6txGXzipFkll8nGduRS7Sk8zybZSBQd5xrMCFrDyG+9yYqi4GoqOaWsakTDnUoGA8kZsamTy46nUAGFA2z39ffggnSlu0CNx0upVro7S3hZ7/SclR50LJowuiaWICaA/z7mmnDDa5HvH88TQQG+hgGAyf4YZWBS7bfzyh0FFLi3N3H7D6oYGPC0LoEG+O9MKBM3ByEQIcWWV0YbAFqXfBDpQcaYUaFLh6fZXqisXArGv1U/QAL0kSyLxkkoSfBW+JDNWG23qYzFYC8yGLJpOqKdb+JgWfAjd8CZx3Tu12AB8D2AfgNUEQ9jDG7mOMcVfJ0wEcYIwdBFAE4A/e+ycC2MwY2wEibA8JgsAtie8GcCdj7DBoJk7G4jFGkJBAs3C164G6bf77oxQh8MQXR1CQkYS7LC+Y/wAAIABJREFUzhuP+ZW5WL6pBh6P0pqyAV1hDiBwTrs6k4fEVPIYENyRDfEehuh1uvHjZduQmWzBw5dPA/Neo8ymBJxSlY8vDjaH/pvhZDxwBo5HBCXS+kYQBCzbWIOLHvsa5/7tSzzz9TFML83GE9fMwoZfnYXHr56Fz+86Hd+YWoJ/rT6CMx5ZjVc31MDlFslP1Qlvbz0Jp1vAlRLtkwCQaKbP4fP9TRCkRhFMgTNwsTlGoITAhaw8yuBjAOd6K5A5AM713hcf0CNCgIMxUtaq10hvU7OOFliBhDF/HC3UxAKDgegpcLmjqWJWHyaBS1VpruFT4KI4AwcQgbPWKc/u4+Ah3nKLer4Ab5Zpo/QZbmisTKpx/Ow8QSRp9g3UuhdMDHra6cStiMAVD4wR4AHdsi2UIllwvZ3UjhmYAed7HQ0KXFdj7Fd6eYVv+pXShCmjhHJ4QmXBtQRECHAUTaGWEKU5ckCAAjesWyghCMIHgiCMEwRhjCAIf/De91tBEFZ4f35DEIQq7zbfFwShz3v/WkEQpgqCMN17+0zAPo8KgjBPEISxgiB8mz8n5jHzavo+BKpwUYgQ2HWiE18dasGNiyuRbDHhqnnlqG61Y/3REBmlBiIDS6q/xc6pQYHjiAEC99L6ary2WUPRNgbx4Af7cLCxC498exry05P6PXb6+EI02/qwN1ScgJgC1+d9jrdA/eneRtzz1i44XB783zcnYf2vzsLT35uD86eU+NSvEdkp+Nt3ZuDdHy5GZX4afvX2Lnzj0a/x5UH9u+0EQcDyTTWYPSoHVUXyRitnTChEg7VX+nMwmckLwdXrj/aJNxMTJZVHxthcxtgJAN8G8CRjbI/3uW0A7geRwE0A7vPeFx/QI0IgEOULgfbj4rMrbhdwYpO/fZIjv6p/VkowuhppMRfpodKEBIoTkIoScNjpRC43AweoM9cAyMAEGBwFzuNUTxI4gZODT0GRU1bDVeBUOH521pICdtrd1OO98t7+j6tpJRZT4JSQ0SyRLLhm7+ejhwLXZyMSqocDZSTBlVup9kmAigOhjHA8bpq35RECHHymrkFFG6XDTqpyBOeaDEQPgiDgWIuMYZESiEUKRCFC4IkvjyAjyYyrF9C56PwpxchKseDVjSIGSAYiD0uyf3HrUhl0HFig0pHAba1pVz0Xaet14v7/7cUv3tiJZ7+W6YyJA6za34gX1lXjhsUVOH184YDHTxtHXW4h3Sh9BC5Agevzur56r1PPrz2OEVnJ+N+PluCmJZUDyGIgppdl47VbF+LfV89Cj9ON657diMc/V1FIVIAt1e040twtaV4SiDO8n83nsm2UyURg+WcQbzECgKLK4yZBEEoFQUgTBCFPEITJAc991ltdHCsIwnOReRsRAG8by6nQb59yeXANO2mBOUqEwAHSi/2uRr8rX6RRPI0InJgayJ0lQ87A5apzoRy0GTiNYd49baEJXFIGDW3LLcDDVeCUEjhBoPeYVUaq1uKfUCB1zQb/NkoiBDjSi+h3FqhcWhW8F77vzoCFGDd6EVXgCumzVtoKqGcGXCQx81rgiheBwony2+VVyc/AdVRTASJYgcuvork4NXNwju5hr74NJfzts0M47+9forMnzHmU4EiBCEcIHG/pxoe76nH1glHITCbDoGSLCZfNKsXHexrQ2hUf4uWQgjmlvwKnZkYokOxl6EPg7A4XrnpqPX7/njqjpk/3NsLh8mB6aRbu+99evLTuuC7HE200dPbi56/vxITiDNx9/gTRbQoykjB1ZFboOThRBc7bQpmUgYONNqw90oprF1b0C8qWA2MMF0wtwad3nopvTivBI58cwNeHWhQ9VwmWb6pFepIZ35ga+jpfkJGE6aVZWBlqDi6eFbhhi+5mOjHpaYdcPI3UMrE5uBoe4B1M4Hjws8RizdZIbWvRQPFU6lnvOD7wMTufVVOgwCk11wBIgTOnRN8BT2sWnBIFDqCFtJwhjK2BXB21uh4pJXD2NlJ4OYFa+EMiYZ/+1m9Tr0aByygCBI9fOQVIPU7JlT/5JabRNv0UuAPUhy5WRPFlwSk8+XMCF+lW43CRWQJMks1/JuRX0WcVnNHD0XLYv10gTBagYII6J0pn97CefxtqOHsiObC9v7M+vB0VjAdGnwFsegYNbTa8v3oNhAhGCDz11VGYExJw4+KKfvdfNa8MTreAN7caZiZRxwAFTmUOHEAFJZ1igtYfbUWfy4NP9zaqKlC8t6MOI7NT8NoPFuLsiUX4v3f34NUN8aXqrjncgm/+82vYHW48etVMJFsk8ngBnD6+AFtr2uXjBPj1mht4AAEKXDqeX3scSeYE2VkzKSSZTXj48mkYW5COHy/fhvpOieuYCrR09eG9HXW4aMYIpCWZQz8B1Ea5vbZDuvgThgK3/mgr3txyIqLzfgaBk4KeEQIcJjNQOtdrVhKEmrV04QtuJUgvIhv1WFDgSqbRrdgcnF2pAued7+tV2EbZ3RLdCAEOH4FTq8B1hG4jBYCyBRQZsedt8cdt9eGpRUoJHCeo/P0mpgGn30MGBfv/f3vnHR5Xde3td6s3S7KaZau4Se5NrhgDNqYECGCMIWDTL4RqQgokcJPcSwppXyCNhN47Mc1wTeg23b1h3OQquUqWZMvq0uzvjz1HGo9mNGfUZjRa7/PomZkzZ0ZHW2dmn7V/a/3W/7XsE51oAkpfNDfzdjEj8dVCwCI598RecKXbjAW+p8bw1jlvt5l3T1Hg7JKaB2jvbUmaWwjkt36u32j/Arj66qAr3hbaz9isJPIzEni9MwKeaTdD5QG2fPICjaWFVERmdkmq7eHKWhatLmbepGwyEk9cCMrv14fJA/vy8ooi74YEQtdwQhuBdipwif07zcV36dYSwhTUNzp4d6O9BYqK6no+217Kd8eZuq1/XlHA6cPT+e83NvaImrgmh+aBD7Zx5RPLSY6L5K2FMxjmo/5r1vB0HJq22wlY/8uG1gFcpSOW19cUc9GELPrGt+/zHhcVwUNXTqKuoYmFL66loYOBztNf7Ka+ycENpwy2/ZozRvRDa7yqkU3hUazffYj31zvjAT8UuD+/t5UHPthm27K/PUgA5w2riXffTl5NHDgDDm860YlRa6PAuatvYOpd2nKiPH6o++p60keaok5PTpSWAucreLEUOrtOlNWl3W9gAqbGIzqxHQGcTQXu1B+bYP7NWz3XI1Ue6Nj/1W7T9OYAzmUVreAqo/x+eK+pzazYa56347bYHMC5BFbH9tkM4Nx6wZVshXQP6ZNgauCgpf7GF80pqUFeA2cXX8p86XZzHnr67PQbbVRRu5/BhmpR4EIIpRQXT8xm1Z5ydne0Fs7ZUiC38DkGqYNsqk3l8LFa36/zk6e+2E1Dk4MbT/NcXzd/ai47S6tYvqvnlNh3FYcrazuurrpQ19jEc1/t5tqnVrSuLYuMaXHoa2itwDU5NC8s38PLK/by0eZDbCiu4MDRGuobHS3fKZ2UPgmwbFsJpw/PYEhaPK+v3WfrNe9tOkijQ3PBOHMc0RHhPHTlJE7NT+Nnr23gjbXBq+weOlbLgse+5u8fbWfexGwW2wjeACbk9CUpNpJPtrQxf4ZFgApzU+CM4cebW45R2+DgmpMHdej48zIS+MO8cazeU87vl7SRkeSD43WNPPvVbr4zKpMh6fb9IEYPSCS9TzQfu9UDOhyaV1buZXdFI/tKK/j0W+fCsk0FbuXuMlbtKef7pw4m0mZ6aXuQAM4bVgCX5L883CZWjVuRS43RkR0mZdO9/s0ibVhLSpQr9dXmA9VdaWGRMSb9ypORiaX02OkD57q/L6pKu9/AxMLfVgKN9SY91E4AFxEN33vOBFovL2h9Md3R3jh2m6ZbAarreR4eAWfea1Sctc/abyEALcqYq5GJXTUxKdekBWptVv0q9piWCx5/j5VC6YcCFxnXdnuHnoRVZ+StDu5IoWf1DVqMTOyqcFIDF3LMLcgiTNFxFc7ZUmBIzTeMDtvDLp3JQ84m251FZW0Dz3+9h3PHZDI4zfN5eN7Y/vSJieClEDIzKSqr5rPt/jv1/Wrxt9z24hq+3NGx2qLGJgevrixi9p+X8cu3NrF0a0nr8Y2MM71rmxpNmpmbAvf+poP8/I1vuPv1jVz/zCoufPALpv/+Y4b94l2ufd58/1REdk6Gze7SKvYcqWbW8HTmFmSxYlcZRWW+G0i/s+EAA1PjGJOV2LwtJjKcx66ezPQhqfzk1fW8vd5u56zuY9m2Es7722dsKD7K/ZeO58+Xjicuyl7qYHiY4rRh6W23E1DqRIUVmtsIPLPqCNMGpzBqQKLn1/rBBeMHcO3Jg3jyi10ssamauvPS8r0cq23k5ln+1d+GhSlmD8/g060lzQrgpv1HueThL/nZaxtRETGclBtPn3BnqqnNFOGHl+4gJT6Ky6Z0YgafBySA80bFXhM4dLa7Y9Ykk/PtamRi1cR5UuDAKHCV+1vyjy26q4WAK5ljvaRQWgqcj+DF3wCu+kj3G5hY+BvAWX+T3b6Bif3hsudNgLPov8wkCMYkpvKgeb4j2AngrB5w7rWLw88zaZ6f/N44p9oO4JznopWy2FhvFifsKnCNNeZ/fqTQ1NL5VOBsBnDHDxr1rTN7tgWSqHhIzG5bgXOvf7PIHGtu7QZwDdXdX4MqdCmZSTHMyEvj9bX7OtxDrTT/Uqp0NBE0kZg1nBeX7+1UFe7F5XuprG3k5pneL85io8K5uCCLdzcepLzKZn11EFPf6OC6p1dy7VMr2VFy3PbrCg9XsuQbcxH8x/9sbVdKaZND8+bafZz5wDJ++toG0hKiePa/pjIjL5VXVhbR5Hq+WAFbfaUJ5NxSrV9YvpcBSTF89tPTefO2GTx29WR+N3csPz5rGFOGmbT9Lw9H+n2MnrBcFWcOy+CigiwA3lrXtgpXeryOL3cc4fxx/Zt7pVnERIbz+DWTmTwohR++so7/fNN5qmZHaGxy8Kf/bOGaJ1eQlhDN27fPYN6kbL/fZ9awdEqP+2gnYJl4WNRVolEUHtVc51aL2hH++7yRFOQm89NFG9jpx/kO5rPyxOe7mD4klQk5/vdsPn1EBpV1jSzdWsKv3t7EBf/4nD1Hqrn/0vEM7p9KSpSDqdnmvN57zPfnaevBSj7acphrpg8iNsp7HWJnIAGcNzq7hYBFZKwJ4twDuLhUz2574D1dygrgOrM5sS/6jzMXw+4XztVlJu0w3McKkJViadeJsqq00wqc/SYp278UymYV0kYNnEXOFPjuA7DzE/joXrOtutQ0N+1ovZYtBa7I/J3ugY1ScPZvjMJVf9yeAyWY8zs6qeX8OO5H7VlzK4G9Lg6UXhS46ARzsWC3lUBHFc1gJC3PcwBXe8yMuzc3wIQMU1d6yEtLEHfqJYUyFJk3MZvi8hpW7O5Y2uGaQ0281nQaANMnT6HRoXl4mZfaTD+pa2ziic93MSMvlXHZbV+czZ+WS32Tw3bqXDDz2Gc7KTx8nDAFf/mgjXYzbvzzkx3ERIRzz7kjWF9UwX++Oej7RU601ry78QDn/PVTfvjKumYV6s3bZnDasHTmT81lX0XNiaqgpUhYrYFcaoR2l1bxeWEp86fmkpMSx4ScZM4a1Y8F03L5wRn53Ha2qalfXR7H2r1+9Ib1wrJtJQxOiyc3NY6clDimDkrh9bX72gxi3/3mIE0OzfnjPC8wxkVF8OS1UxifncQPXl7H6j2BT9H9xZvf8K+lO7h8Sg5v3jaDvIz2ZZWcZqedQGRsqxq4ahVHVnIcZ47svOvOqIgw/rlgIpHhilueX0N1faPt1765bh8Hj9X6rb5ZnJKfRmS44sbnVvH0l7u5YtpAPv7JLOZNykZFRENjHVOzY2nUYTz0+R6f7/fIsh3ERoZz9fSuMXNyRQI4b1Ts6fz6N4uB042BRb2z/mDPl0Z986YO+ArguluBg9Z1cNVH7Jl3+KPA1VcZRSZgClyOOc46mytCzQqcjRRKVyZeBVO+D1/+Aza82tInsKP1WrYDOC/BWc5UGHmBue/PYoZrLzjrb0nM8v0663dU7DU94FSY06zDC/Fp9hW4o8WhF8Cl5juVSrcLFKtJt7cUSvDPyKRBUihDke+MziQhOoLXVncsjXLN3gqe0BfQNOJCMkbPYm5BFi8s39MpKtwba/ZxuLKOW2a28T3gZERmIhNyknlpxd4ebWZSVFbNPz7ezjmjM7nptKG8s+EAm/Yf9fm63aVVvLVuH1eelMsNpw4hPyOB//feVtsueA8t28EtL6yhSWseXFDAkh+cylmj+jUrU2ePyiQ1PurENEpLgbPmGZcUypdW7CU8TPE9by6Fybk0jLmMFZFT+dfSjqXd1jY08dXOI809zgDmTsxiZ0kV64u9j9076/czND2eEZneg6CE6AieuGYKWcmxfP/Z1ew50sG60Q7w71VFvLyyiFtmDeUP88Z1SOGx1U7ATYGrqCijwhHDVdMH2m4dYJcBybH87fICth2u5BdvfGPrM+xwaB5etoNR/RM5Lb9914kJ0RHMLchiUm5f3rptBr+5aAxJcU5VOCIGmuqID2ukKTyaRauL2Vfh3TGzuLyaxev3M39qbrvNXfxBAjhPOBwmtayL7JAZOMOkGxSvMspA+S7IPcn7/n0HG0MKdyOTSiuA60ZjhuYAzm31vqbMnlIW3ccUx9oxULDUlUC4UEJLYHPM5oqupSr6G8ABnPN7c14svh22vWe2dYsCV9ziQOmJs34DQ2ebdEq79MlsCayaAzgbf0tzL7gio8AlD2zb9Sk+w14NXEONCQq9pRT2VNLyTQ2sexB7xEsLAVf6jYHDm0/s1+cNUeBCktiocM4bm8mSjQf8WvF2Z+3ecpL7DyX88ucgNpmFp+f5pcLtKDnOlztK+fDbQ7y1bh8vrdjL45/t5O8fbecfHxcyJiuRGXn2sjAWTM2l8PBxVu3puKITCLTW3Lt4E2FK8T8XjOL7pw0hMSaC+9/3rcL9a2khEeFhfP+0IYSHKe76znB2llbx6irfAfq6ogoeeH8b543N5P0fnsb54wYQFnbignJURBiXTMrmw82HW4LzZgWu7ITHdY1N/Ht1MWeN7Ee/RC/f4eGRRF7yKLNnTOeDbw+x9WCl5/1ssGJXGbUNDmYOb7lWOG9sf6IiwnjDS53noWO1rNhdxgXjB7RKn3Snb3wUT147BYfWXPf0yrbt97uIzQeO8cu3vjF1eWd5ydbyk9Od7QQqqr2kHbvVwBUdPEw1sVw2uZO9IZycNiydO87I5/W1+3j6y90+9/9g8yF2llRx86yhPv+HbfGnS8az6JaTW6v8TgWOxhoios0c+EgbNb6Pf2aawN9wqn0nzI4gAZwnKg+YJrhdkUIJRtlAGeWtuf7tZO/7R0SZXljuAdzxQyaw684Uw9i+xmzCvQ6u+oi91EG75hoAVVZvuQDWwIH9OrhmBc6PFEqL8Ei49Bnzty79ndnWGQpcdRvj3FBjguS20iNTBsNVb/iXppuQ0ZI66U9D8phkiOpjFk9Ktpk+Uz5/j40UyiOFgA69AM5SJ92NTI4UGvUyxbNjH2AUuMZa720IXGmQNgKhysUTs6mqb+L9TYd87+yBxiYHG4qPUpDbsmg1KC2+RYWr9K7Caa355yeFnHH/MhY8tpwbnl3FHS+v457XN/Lb/9vMAx9so6yqnh+dOcz2xdn54/uTEN1+M5Mjx+uobbCxqNFFvP/tIT7acpgfnTmMAcmxJMVGcvOsoXy85TCr2kh1LS6v5vU1+5g/JYeMPiZgOmtUPyYN7MtfP9xGTb33v+l4XSN3vLyWjD7R/H7uuDaVlcum5NDk0PzbUm0txa36RAXuP98cpKyqngXTfF9DXXvyIOKiwnloqQejNpss21ZCVEQYJw1uuRZKio3krJH9eHvDAY8W9Us2HkBrvKZPujM4LZ5Hr5pMcVkNNz2/yjhpdoDPthtTGDs1qJW1Ddz6whoSYyL52/wJnaZ+zRyeYdoJeGum7aLAVVTXc6ziCFFxSV2qLv1gdj5njerHr97+ljfbSIfWWvPQ0h3kpMRy3pguEjHCWxp5h0fFMW9iNi+vLPKYXVBeVc8rK4u4cMIABiS3s3+vn0gA5wmrF1VXpVDGJEHmGNP7be/X5uLI6rHmjbRhHlIoD5qL2E7qoWKb/uM8pFCW2w8k7QZw1c4vlUCamID9Orj2plBaJKTD5c87J0HV8dTYuBRTXN7kZbXwqPPLsbOdVhP6najARcTYGxOlTDBZvtsEId5qQi3i0+0pcCU+6ul6KlZA6v69ULrdLD5FRHt/bbMTpYcWFq44HGJiEsJMHZRCdt/YdjfB3nKwkpqGJiYOPPHzbalwj3hR4bTW/P7dLfy/97Zy0YQBvHzjSSxeOIMPfzyTr+6Zzfr/PZvt953L5t+cwxl+1NrERUVwUcEA/m/DAb9VkqVbD3Pqnz7huqdWdtjYpT1U1TXyq8WbGJHZh2tdDCKuPXkQaQnR/Ok976YkDy/bgVJwk4vRi1KKu88dweHKOp76cpfX33vv4k0UlVXzl8smtKSOeWFIegInDUnh5ZXOwKNZgXPOfc7HLyzfS25KHKfk+Z67+8ZHccW0XBav38/eI75dIz2xbFsJJw1JbZVSOLcgi7KqepZ5SBN8Z8MBRmT2IS/DvlHd1MEp/PGSsXy9s4z/fmNju1J191fUcPNzq7nqiRXc8/pGbnxuNcdqvZ+rWmt+9toG9pZV84/5Bc0BemcwISeZ5LhI72mUkbFGgQJeWVlEnK4hNbVrBYOwMMU/5hcYpfHf6/los+fFpeW7ylhXVMGNpw3t9HTOZlwbeUfEcMusoTQ5NI9+2vp77ZmvdlPT0NSm2VJnIwGcJ6wWAl2VQgkmXa5oJez6FLInGwWmLdLyoWxHi1MhmIvk7mri7UrmONP6wLU2zG4NHJj97JiYVDkDuECZmPTpb5QMfwI4Fd4xq/oBBXDJkzD9Nt/nhC+a6w29NE0/6lyo6PQALsMYn9QdNwpc4gD77o9JOWZRo6nOngJXfcR3GmCpjXq6nkhitklxOeK2cn1ke9v1b2CCWRXuuw6u0ZnvLwpcSBIWZnrCfV5YyoGj3ms7vGGZTxS4ub8NSovnoglZPP91axWuyaG5+7WNPPrpTq6ZPpAHvjeBk4YYk5K8jAT6Jxnlqb39ky6fkktdo4P7lnxrW03796oirn9mFfHREXy18wgvr+z+Bs5//2g7+4/W8tuLxpzwt8dFRXD77DxW7CrzqJQcPFrLqyuLuWRSTquV/ymDUjhjRAYPLd3hMU3unQ37WbS6mFtn5TFtiL15dv7UXIrKavhyxxGPAdz2Q5Ws2FXGgmm5rdIwvXHDqUOICAvj4U/9r4UrLq+m8PDxE+rfLGYOTyclPoo33JScfRU1rN5TzgXj/e9BN7cgmx+emc+i1cX88xP7qmF9o4OHlu7gjPuXsXTbYe48exi/PH8Un2w9zEUPfkHhYc8ppE9/uZslGw9y13eG2/4f2SU8THFqfhvtBCKioaGGxiYHz361h7SoehL6+O/06C8xkeE8ds1kRg9I5NYX1vD1ziOt9nl42Q7SEqK4tB0OnLaxFEhnj8OBqfHMGT+AF5bv5cjxlp6I1fWNPP3lbs4cmWGrD19nIQGcJyq66MLWldzp5uLo8Ldtp09apA2DpvqW4BJM/Vx31r9ZZI4FdMvFX0OtMTqw677YUxS48AjTZNSfAC62b8et6kd8F75zX8feA3wbxjT3gOvkL0DrnDx+yChw/jRqTc6BOmfRuS/FLD7DtBqobv3lfgIlNurpeiJhYcZp0lWBczjM4oqvdNHIGLOPrwCu3rkiLgpcyDJvYhZaw5tr/e91tWZvBel9osnu2zpl6PbZRoV71EWFq2tsYuGLa3hlVRE/OCOfey8cbfsi3y5jspK44ZTBvLqqmDkPfsHmNmzStdb8/aPt3LVoAycPTeWjn8xk+pBUfrdkc7sC2vay9WAlT3y+i8sm5zB5UOt5dP7UXLL7xvL/PKhwj3y6gyatudWLC99PzxnB8brGVkYh+ypquOf1jUzISeaOM+2nl39ndCbJcZEmTbWViUksL67YS2S44hI/Lqz7JcZwyeRsFq0q5pCf5jfLthn1yFMAFxkexgXj+vPB5kMcrWlRuf5vgznXzx/XvjrzO87IZ25BFn9+fxuLbfSI+3JHKef9/TP++J8tzMhL44MfzWTh7HyuP2UwL94wjWO1Dcx58ItWrQpW7ynnvv/bzJkj+3HjqW2kxHeA04ebdgKb9nv4nDgVqA83H2ZfRQ1pkfWmd203kBAdwdPXTSUnJY4bnlnFRhczms0HjrF0awnXnjyImMgutOq3FDiXJvW3np5HrdMd1+KVlUVUVDdwSzudMNuLBHCeKN9jLkK78oJvoEvQ1paBiYWVTua62h4oBc5K97TSKC01za8Azosq5EpVqclBjurkXnz+4E8rgeoy/1oIdDVWPzpvAVxFkVGm7PRo84cElx5tx/b718/Ote7UWw+45t+T3vJ72qJ0u281r6eSmndiDVzlfpPyaEdt7DcGDvpIoWxwOq6JAheyDEyNZ/LAvry2ptjvlLC1e8spyEn2WKPWrMI5a+Gq6xu54ZlVvPvNQX55/ih+fJb92jZ/+cX5o3jquimUVdcz58EvePTTHa0UhsYmB//9xkYe+GAbF0/M4olrppAYE8kf5o2l0eHg5zad8DqKw6H5xZsb6RMTwd3njvC4T1REGD88cxgb9x3lvU0trQFKKut4cfle5hZkkZPi+TM6PLMPFxdk8/SXu9nvdNBrcmh+9PI6HA7N3y6f4JfaGRMZzryJ2bz/7UHKG5wXz845ppYoXltdzDlj+pOW0EYKtwduPm0ojQ4Hj3/mXwuKZVtLyO4by9B0z4tMcydmU9/o4F2XJtHvbDjAuOwkBqa2b2FKKcUf5o1l6qAU7vz3el5fU8yn20pa/SzbVsIdL69lwWNu4SOtAAAgAElEQVTLqW1o4olrJvP4NZNP+F9NG5LK27efQl6/Ptz8/Br+9J8tNDk0ZVX1LHxxDf2TY7j/0vGdvtBh0WY7gYgYaKzh6S93kZUcS4yjqmMZRn6SEh/Fc9dPJSk2kmueWkHhYZP19fCyHcRHhXPVSYO69gCaTUxqmxcr8jISOG9sf579ag8V1fU0NDl4/LNdTBnUl0kDu/f6TwI4T3RlCwGLhAxzkaXCIXuK7/2b612cRiaOJlP/01Gji/aQmGWCMCuAsxwl7aY6xqXYc6G0mngHsvmyP828LQUuWGhW4LyMtWWt39FUTXes2r3jB/3vv2ap3gmZpla0Laxm3m3VwTma7NXT9VTS8s2CU6MzPcpS4+wYtvQbbdJoa9uwKG9W4CSAC2XmTcqm8PBxNrRhue7OkeN17D5S3ar+zZWFs/Oob3Rw/3vbuPLx5XxRWMqfLhnH9ad0vUvb6cMzeO+Hp3H6iHR+t2QLCx7/utkCvLq+kRufW81LK4q47fSh3H/peKIizOXQwNR47jx7OB9vOWxLXekoi9YUs3J3OfecO7JNc4i5BVnkZSTw5/e3NTfTfvzznTQ0Objt9LYXbH50Vj7olp5yDy0tZMXuMn49Z0y7gpj5U3NoaNK8u8W5EOucY5burORYbSNX2DAvcSc3NY4Lnelpdpux1zc6+KKwlJnD0r0uBozPTmJIWnxzf8A9R6rYUHy03eqbRXREOI9cNYms5Fh+/Op6rn5yRaufa55cwbsbD/KD2Xl8+OOZXus5+yfF8upNJ3H5lBz+tXQH1z29kjteXsuRqnoeumKSz9rEjpCWEM247CSWOpXMmvom1hVV8NKKvazZX0NJ+TG+3lnGVSflouqOd2sAB2Zsnr9hGmEKrnpiOV/vPMI7Gw6wYFpul44LYII2R4OZByNbsgxun53H8bpGnvpiN2+v38++ippurX1rPrxu/409gYo9/tmmt5cJC4zbXrQNhSkuxTgUWgFc9RGTPtadPeAslDJ1cFYrASuFzXYNXLJZ2W+sa9tooaokcOmTFknZ8O1bJjXNl1lMTXnnq1kdoblpurcUyjZ6wHUEa1Hh8BZTy+bPmFgKnC/1DVyUvjacKMt3m2MI1QAuNd80fS/fZVRGOz3gLPqNMbeHvjW9KT1h9aqUPnAhzXfH9ed/F2/i9TXFjM+xV+OyrshcvE/M9R7ADU6L56KCLF5ZVURUeBj/umIi54zpvn6MKfFRPHzlJP69uphfLd7EOX/9lLvPHcGrK4vYuO8ov71oDFee1Hqx9roZg3lnwwF+9fa3nJKXRqqfapJdyqvq+f2SzUwe2NdnymF4mOInZw3jlhfW8ObafcwekcFzX+3hgvEDGJzW9uczu28cV00fyFNf7GLK4BT+8uF2Lhg/gIsn2ujP6YG8jD5MGdSXResPsACaF2Tf+OYIQ9NTmTa4fUrELbPyeHPdfp7+cjc/smGVv3pPOVX1TR7TJy2UUlw80aQ7FpVV884Go8R916b7ZFv0jY/indtPYctB72m6WclxZCb5zuaKjgjnD/PGMT4nmf99axP1TQ5+N3csY7J8LGR2ArOGpfPgJ4Wc+cAydpYcxxKr/xjdyNDwOm6aOYSrJ6fBUm3verWTGZwWz7P/NY3LHv2KBY99TXiY4vpTuial9AQinAsqtRWmXMHJiMxEzh7Vj6e+2EV6n2iG9+vD6cO7PxtOFDh3mhqNO19XtRBw5dSfwMWP2N/f1Ymy0plGEYgADkwd3KFvjcNhjZ8KnC9zDYuq0sC1ELBIyjYrMHbcDmsqglSBayuA64IC4NgUoyzvX2se+xPAWQGlHcdIqz9gW/8ba8EjVFMo05wr79b3Qul2k3JsR5m340RppVCKAhfSJMZEcvaofixev9+2PfqaveVEhCnG+rjA/NGZw5g2OIUnr53SrcGbhVKK703O4d07TmN4vz78/I1v2HqokkeumuwxeAMTLP3pknFU1jbwq7e/7fRjcjg0H357iKufXMGx2kZ+O3eMrRS5c8ZkMjYrib98uI2HP91BdX0TC32obxa3nZ5HfFQEP120gczEGH570ZgOpbDOn5pLYbnTJMY5x6zZV8uCaQPb/b7DM/tw1qh+PP3lbo7X+e5NuGxbCZHhipN9uF3OmWAC1bfW7eOdDQeYmJtMVidZvcdHRzBpYIrXHzvBmyvzp+ay6Jbp/OHiscyf2oU+DC7MKchieGYig9PiuX12Pg9fOYnPfno63zspj6QIB/ecO5I4h7MmtJsVOItRAxJ56topzn6EOX6Pa7torvGsOEGBA7h9dj7HahvZUVLFTTOHdFmKa1tIAOfOsX1mRburUyjbQ1p+ywWpVfcTiBRKgP7jjbJRuq1FgbNdA2cpQz7SKKtLg0CBc36BVthIo6wpa18PuK4iOtHUuHkK4BwO50JFF0wQYWFGHbMCOH9MTBIyoOAqGDPP974xSRAe1XYNXHMLgRBW4KClDu7IdrNSaOcCKnGA6b3XlpGJlUIpNXAhz7xJ2ZRXN/DxFhuLVcCaPRWM7J/YyrrdnZyUOF65aTqn5Af2uzw31RzHHy4ey6KbT+asUW0vfg7r14eFp+ezeP1+Pvy2fX3y3GlocrBodTHf+eun3PDsKsqq6nnge+MZkWnPGEIpxZ3fGU5xeQ2PLNvJuWMyybfpepcSH8XC2XlEhiv+evkEkmI7ln523tj+RMQ4lT/nHNMYEcO8dqp6FrfOGsrRmgZeXL7H575Ltx5m8sAUEqLbTibLSYlj6uAUnvpiN5sPHGuX+2R3Mi47mcun5nZZjag7Q9MTePeOU3ns6sn86KxhnDMmk5yUOFSkSyPvOqdLZjeZmHhi8qAUvrr7DH4zZ3T3/EIrQ6y+siWYczI2O4kzR2aQmxIXsPNJAjh3mlsIdIMC5y9pw0ywVHWkpVFyIExMwOlEiUmjrPazgbUvZcii6khwKHDguw6usd5Y5weTAhcWZi7QPY3z8UNGWewKBQ7MeWkpY/6YmCgFcx70ntLnvm98Rku7CU+UbjMqdWzXWx8HhJhE8/eVOlMnSwvtpU+CGb9+Y9oO4BrEhbK3cGpeGul9onndRk+4JodmfXEFE3N71ucqPExx+dRc22lpt8wayojMPvz8zY1t9uryRVVdI098vouZf/qEO/+9nvAwxV8vm8DSu2Y1q0N2OS0/janOFMWFs/1rjXLTzKGs+vlZTPHgdOkvMZHhXFgwkCat0LUmm+aMsQNJjutYk+eC3L7MyEvlsc92tdkG4tCxWrYcrGTmcO/pk65cXJDFkap6lDLBp2ADqwbM0WSCGAisqRwmbbXL+r654xq0eVjEfHDBRN6+/ZR2tzzpKBLAuVPeDT3g2kuay2r7ceeKYKBSKFPzzcl9YIMJKqP6tOQL+8JOANdQY9K34gPUA87CbjNv5wQWdIGCt5YNVkCa1EULFc3tLTqhIXmbv8dHM+/SbaGrvlmk5pvvhIYa83+1Y2Bi0W+0aWXi8JI2Z9XASQAX8kSEh3HRhAF8vOUwh31YuW89WEl1fRMFbdS/hQJREWH8cd44Sirr+P2SLX6/XmvNk5/vYsYfP+Y373xLdkocT107hXfvOJWLCrLadeGnlOLPl4znwQUFjB7gf31UZxo/XD4tl1qiUNpBow5j/kmdY+Rw26w8SirruHfxJq8pvVb7gFk2A7hzx/YnKiKMqYNS6JcYYi1lugpLgWqsdVHgApNCGRBcPRo8uNLHRIZ3WMnuCBLAuVOx16SddZUy0RFcnSgrD0F0Uqu83G4jPAIyRhknyho/7fOtfdtyomxu4h1gBS422aQM+ArgrCApmBQ4MMfjaZybA7guVODABG+d7XLpSnyG9xRKrY1JUKjWv1mk5Znat7KdgPavYXm/0UY5rtjt+fkGSaHsTVx10iA08PCytq3c1xaZ77u2DExChfE5ydxw6hBeWrGXz7a3YZjkRlVdIwtfWsuv3/mWcdnJvHbLybx603ROH5HR4dS43NQ4zu8EE46OMiIzkcYwc5Fbr6I7TZGdPjSVm04bwssri1jw2NceFxSWbS2hX6IxkLBDUmwkj1w1iV/PGdMpx9grsK4vG3prAOcStEUE6Fq7DSSAc6dij7HJ78qLzvaSPNDU/JRuMwpcnwCpbxb9x5kArqrUvwDOjgJX5Zwo4+2trnUpdloJVPvZC6+7iEvxosB1URNvC0t18yd9sl2/J73lXHHn+CHTFLw3KHA1ZVC03Dz2R4HLmWput77r+XlR4HoVualxXFyQxQvL97Spwq3ZU0FqfBQ5KcF3UdMV/OjMYQxJi+e6p1byuyWbqfJhsLG7tIqL//Ul7248wN3njuCZ66YwqY12Cz2ZiGizuKOiYjqtZkspxT3njeTv8wvYtP8Y5//jc1bvaZnHGpscfLa9pM32AZ44fXgGwzN7UQDSUU5Q4EwPtl4VwIW3rcAFGgng3KnYG5zpkwBh4WZ1vdSZQhmo9EmLzLGmh9TBDf6Zd0QlQFhE2wGcZYwSaBMTMApOW059ENwKnKdxrigyJiAxXVSQbJ2b/hiYtIf4DBPAeUoBDHUDEwsrYNv6H3PrjwKXMRKyp8LKJzyPYUO1cRQN71hdSyiglDpHKbVVKVWolLrbw/MDlVIfKaU2KKWWKqWy3Z5PVErtU0o96LJtqfM91zl/AlTU3MLC2Xk0OnSbKtzaonIKcvt2m8lCoImNCmfRLSczb2I2j366kzMfWMaSjQc8Nvr+eMshLnjwcw5V1vLMf03l5plDQ3qc4uJMTVRMTOcv8lw4fgCv33oyMZHhXP7oVzz/9R60NvWXx2obmRUA6/ZehaU6SQplKxOTYEACOHfK9wSngYlFWn4QBXDjzW1Vif0WAmDME2JT2nahbE6hDHANHMCAAhPYVx3xvk9QB3Ae2jUcLe6aHnAWfbpLgcsAR2NLDaIrod5CwMIK2HYuNdkD/qplU26Ash2wa1nr5+qrzfuF8AWoHZRS4cA/gXOBUcB8pdQot93+DDyrtR4H/Br4vdvzvwE8DDJXaK0nOH/sWUB2IQNT49tU4Sqq69lZUsXEgUFW79vFpMRH8cdLxvHaLSfTNy6KW19Yw9VPrmBXqVGpHQ7N3z7czvXPrCI3JY63F57CqflBkEHSxaioWOdt16RZj+yfyNsLT2FGXhq/ePMbfvbaBt7fdIjwMMUMH+0DhA5yggLn7HXXqwI4VxOT4Ms2kADOlcY6qDwQnC0ELNKGmaa9xw4EPoDrN9rUC4L/qYPelCGLamcAFwwK3IACc2vZ4nsimAO4uqOmv6ErXdXE26I5hbKrFTjnBZKnOrjSbcZcp0+IO44lD4SwSNPWwx/1zWLUHLOgsuqJ1s81VEn9m2EqUKi13qm1rgdeBua47TMK+Mh5/xPX55VSk4B+wPvdcKwdpi0Vbu1es1hSkBNk33XdxKSBfVm8cAb3XjCKdXsr+M5fPuXP723lxudW85cPtzF3Qhav3XIyOSm95HNjqTRdqFAkxUXyxDVTuH12Hq+uKuaRT3dSkJMcUAOJXoF7DVx41ImqVKgjClwP4mgxoINcgRsG2gGNNYGvgYuKa7lg9Fcp81abZVFVai5KA9hzpJn+TqWxzQCuzKSaBcPxumIFlLVHT9zeVU28LfoOMmmydhpyd4S2mnmXbIX0YaGvHoVHQMoQc9+f+jeLyBiYeBVsWQLH9p/4XH21NPE2ZAGuhbDFzm2urAesBoZzgT5KqVSlVBhwP3CXl/d+ypk++UvlJddOKXWjUmqVUmpVSYl9I432MjA1nrmWCld5ogq3dm854WGK8Tn+OyCGChHhYVw7YzAf3TmT747rz4OfFLJ062HuvWAU939vPDGRbffGCyms2qAuVijCwxQ/OXs4j141iaTYSOZMCLyJS8jjqsDVHw94C4Fux0cbgUAjAZwr5bvNbbDWwMGJF2iBVuAAMseZW3+Vp9i+Lf3jPGE18Q6Gi++YJBOo+lLgYvsGx/G60mwY45KuWnvMBHRd0cTbok8m/HAjjPhu1/0OaHG79KbAdXUAGSxY3wt2e8C5M+k6szC0+pkTtzdUQ6QYmACePtjuBVB3AjOVUmuBmcA+oBG4FViitfbkhHSF1noscKrz5ypPv1xr/ajWerLWenJ6evek5S083ahwj7ipcGv2VjAisw9xUW03T+4NZPSJ4S+XTeD1W0/mjVtncO2MwSFd7+aRblDgXDl7dCZrf3kWV54UxNdpoYL1P7Vq4HpT+iT4bCMQaCSAc6Vir7kNZgUuNdgCOGdDb38VOF8plMHQxNuVARPtBXDBhmUu4zrWXe1AaZE4oOsD2nhnAOfuRFl71KRDt0eR6olYSnh7UigBUgZD3pmw+mlocmlYXF8lCpyhGHBd8cgGTpArtdb7tdYXa60LgJ87tx0FpgMLlVK7MXVyVyul/uB8fp/zthJ4EZOqGRQMSjMq3PNft6hwTQ7NuqIKCnpYA++uZmJuX8Zm91JFspsUOFfCwlTvC5QDQasALsgyjLqaE1Ioe2gNnA33rWil1CvO55crpQY5tw9SStW4OGw93LmH38lU7DFpe11dt9MRohNanP2CIYCzbMj9DXp9BnAlwVH/ZjGgACr3Q+VBz88HbQDnoWVDVzfx7k5i+5rUVXcFrnS7uQ11AxOLrInmu6vf6Pa/x5Tr4fhB2LqkZVtDdVCmjgSAlUC+UmqwUioKuBxY7LqDUirNmS4JcA/wJIDW+gqtda7WehBGpXtWa323UipCKZXmfG0kcD7gw+62e3FX4QoPH+d4XWOv6P8m2KSbFTihGxEFruV+T1TgbLpvXQ+Ua63zgL8Af3R5boeLw9bNnXTcXUPFXqNKhAV5/rqlKvTJDOxxAAw8GW79GrIn+/e62L7GIKGxzvPzVgplsODLyKTaz2bm3UWsc6XcYwAXhM3q/SUszNTBudfANbcQ6CUB3MgL4UffdMz1M/9sY2yz8vGWbfVV0gMO0Fo3AguB94DNwKta601KqV8rpS507jYL2KqU2oYxLLnPx9tGA+8ppTYA6zApl491xfG3l0Fp8Vw0oUWFW7PXfI8USAAnWFjKmyz0hB5W0GKZmET34hq4HqrA2XHfmgNYxROLgDO8FWMHNcHeQsAiY6T5sgwWxSdjpP+v8dXMO9hSKPuPM46b3gK4morg+X+44mmcK4qMWhMMCm5nkJAOx91SKEu3GcesvoMCckjdjlIdX9AJC4dJ18KuT6HE2YKhXlwoLbTWS7TWw7TWQ7XW9zm3/Y/WerHz/iKtdb5znxu01q1Wp7TWT2utFzrvV2mtJ2mtx2mtR2ut79BaN3XvX+Wb252OlI8u28naveWkxEcxKFXOCcFJcwAXfAqF0EF6uwIXAo287bhvNe/jXKk8ClhFUYOVUmuVUsuUUqd6+gXd7bDlkaP74NCmntH099Q74ao3g88wwx/iPNRmgblg/HYx1FdCfBD0gLOIiof0EW0EcEGaQhmTBCijEFocLYakLKNehQLxGa0VuNJtkDLUODQK9pl4tQnuVz1pHjeIC2Vvp1mFW76Hz7eXUpCTLPVHQgvWRX4QKhRCB+ntAVxYmFkIhqA8v+1cwdlx3/K2zwEg11nU/WPgRaVUqyrIQDhsteKDXwIaTl4YmN/vDwnpkDst0EfRMaxgp7rMtAxY+zy8NB/+NARevQpikmHgjMAeozsDCkwAp91O/6YGE3AGYwAXFm6COPcUyq7sAdfdJGS0VuBKtvYeA5POJCEDRl0I6140iyn11b3POlpoxcLZedQ3Oth/tFYMTIQTEQUudHEN4HpjGwFoGYMe2sjbp/uW6z5KqQggCSjTWtdprY8AaK1XAzuA4JO4dn8O37wGM+7oPSlXgcYKdt66Df6cb24PbjQpXNe8DXcVmvq6YGJAgTFXObbvxO3B2sTbwr3n3tHi0ArgrBo4K7BurDPN7nuLgUlnM+UG0/x94yIxMREAGJwWz0UFJvFGDEyEExAFLnSxTDzqjpu5oLe5UELLGARhAGcnv6jZfQtTZH05sMBtn8XANcBXwCXAx1prrZRKxwRyTUqpIUA+sJNgoqkRlvzUXNDO+GGgj6b3kJxrlKGoeDjtLtMvLHNccKeFWkYm+9acaAAS7AGcq+NnU4Ox1+/KHnDdTUIGNNWb1gGxyXBkh+lp1lsMTDqb3OmQMQq+/hegJYVSAOBn54ygX2IMkwcFoVmTEDhEgQtdlDIBenWpedzbUijB1MGpcAiPDPSRtMJnAKe1blRKWe5b4cCTlvsWsMpZwP0E8JxSqhAowwR5AKcBv1ZKNQJNwM1a67LWvyWArHoCDm+C7z0nFyrdSWxf+Nme4A7Y3Ok3BsIiTBrlqAtbtveEAM6qgTu23wQ3oeBAaeHaCy42GUqdDpTpwSf29wiUgsn/BUvuNI+lkbcA9EuM4WfnjAj0YQjBRqS0EQhpIqKNqRz0zgAuIjoo1Tewp8ChtV4CLHHb9j8u92uBSz287jXgtQ4eY9dRVQqf3AdDZsHICwJ9NL2PnhS8gVlhzBjV2sikJwRwR3aY+80tBEJJgXPWzR4/bOreSrcD6sSm94J/jLsMPrzXWfcgC1uCIHihuUZIvidCkojYFpOw3tZGAMz5HaSLEyFiQ9dOPvqVKdQ/9089L5gQAoMnIxNL3QrGPnDgTKF0HuPRYnMbSgFcswLnnGRKtpoUUQk82k9MIoz7nrkvF2aCIHhDUihDm4hok90CosAFGb03gNu3GtY8B9NuFrMDwT4DCqC2Asp3t2wLegUuxdSHOZpMDzgwbQRChQQrgHPm6Zdu7RntQIKdqTcZN9jUvEAfiSAIwYqYmIQ2kbEtc2uvNDERBS64cDhgyV3GvW7mzwJ9NEJPImuiud2/pmVbTbkpcg3WLzcrsKw9alIo49ODdkWpXcSlmibrxw+bz3ZpoRiYdAYZI+Bnu00Te0EQBE/Ep5nbuCDq2yp0HhHRJpUeemkbgeigVZdDr8utwwFPng3ZU2D0XHPrnh65/kWjwF30sEkVEgS7pI80rkT718KYeWZbTbkxzwjWNFwrgKspD70ecGB63cWlmhTKo3uhsUYMTDqLYD2nBUEIDvqPh+9/0uLSLIQWrspqb0yhnHRtSwAbZIReAFdbYWpiVj5ubLCTcmD0RSaYGzDRqBAf3gvZU02hviD4Q0QUZI6B/etattWUmTTFYOWEAK4Y0kPQSS7e2cy7ZJt5LAqcIAhC92Blpgihh9UHDXpnADf6okAfgVdCL4CLS4H5L5pAbeu78M3r8PXD8OU/IHkgJA4w+bxXLIKw3plBKnSQAQWw/hWj9oaFORW4IK1/g5Zjqy4zNXD5Zwf2eLqC+DSjwJU6AzipaxUEQRCEjuFa/9UbA7ggJnQjmJgkGH85XPEq3LUd5vzTFOMXrYApN8CACYE+QqGnMqAA6ivhSKF53FMCuCOFJr0wlHrAWSRkmBq40q0Qlxa8jqCCIAiC0FOIdGkTERYe2GMRTiD0FDhPxPaFgivNT32VuCUJHWOAZWSy1tRa1ZSb2rhgxQpmDm4wt6FWAwcmhbLKmUIpDpSCIAiC0HEsBU7Ut6AjdBU4b0TFS+qk0DHShpnVKKuhd01FcCs+MUnm9uBGcxuSClw6NFSbv1EMTARBEASh40gAF7RIJCMI/hIeAZnjTADX1AB1x4I7hTIs3ARxJVvM4+TcwB5PV2A1826oEgMTQRAEQegMrACuN7YQCHIkgBOE9jCgwKQkWg0ugzmAA3N8jkajHAb7sbYHq5k3iAInCIIgCJ1BpChwwYoEcILQHrImmpS9oq/N42APiqzjS8oJzd5e8ekt90WBEwRBEISO05xCKT2Tgw0J4AShPVhNS3d8bG57TAAXgvVv0KLARcZDYlZgj0UQBEEQQgGrD5wocEGHBHCC0B5ShkJUHyjsKQGc02QlOQQdKKFFgUvLE5MiQRAEQegMLNf2aKmBCzbkSkcQ2kNYmOkleKzYPA76AC7EFbjwSBPEBXM7B0EQBEHoSYgCF7T0jj5wgtAVDCiA3Z+Z+z0mgAtBB0qLy16AxAGBPgpBEARBCA0iLQVOArhgQxQ4QWgvVh2cCm/ptRashLoCB5A7LXRTRIWAo5Q6Rym1VSlVqJS628PzA5VSHymlNiilliqlst2eT1RK7VNKPeiybZJSaqPzPf+uVCg6DAmC0GOxFLgoCeCCDQngBKG9WAFcbHLwOzumDTO57GlisS8I/qKUCgf+CZwLjALmK6VGue32Z+BZrfU44NfA792e/w2wzG3bQ8CNQL7z55xOPnRBEIT2EyEKXLAiAZwgtJe+gyAmOfjTJwHyzoCf7oCEdN/7CoLgzlSgUGu9U2tdD7wMzHHbZxTwkfP+J67PK6UmAf2A91229QcStdZfaa018CxwUdf9CYIgCH4iNXBBiwRwgtBelILBp0LfwYE+Et8oBVHxgT4KQeipZAFFLo+LndtcWQ/Mc96fC/RRSqUqpcKA+4G7PLxnsY/3FARBCByJWYCC5BCun++hiImJIHSEuY+A1oE+CkEQuhZPOdLuH/w7gQeVUtcCnwL7gEbgVmCJ1rrIrcTNznuaHZW6EZNqSW6uXEgJgtBNZIyAuwohPi3QRyK4IQGcIHQEUbUEoTdQDLg65GQD+1130FrvBy4GUEolAPO01keVUtOBU5VStwIJQJRS6jjwN+f7eH1Pl/d+FHgUYPLkybJiJAhC9yHBW1AiAZwgCIIgtM1KIF8pNRijrF0OLHDdQSmVBpRprR3APcCTAFrrK1z2uRaYrLW+2/m4Uil1ErAcuBr4R9f/KYIgCEJPR2rgBEEQBKENtNaNwELgPWAz8KrWepNS6tdKqQudu80CtiqltmEMS+6z8da3AI8DhcAO4N3OPnZBEAQh9BAFThAEQRB8oLVeAixx2/Y/LvcXAYt8vMfTwNMuj1cBYzrzOAVBEITQRxQ4QRAEQRAEQRCEHoKtAE4pdY5SaqtSqlApdbeH56OVUq84n1+ulBrk8tw9zjkC9WsAAAanSURBVO1blVLf6bxDFwRBEARBEARB6F34DOCUUuHAP4FzMY1K5yulRrntdj1QrrXOA/4C/NH52lGYYu/RwDnAv5zvJwiCIAiCIAiCIPiJHQVuKlCotd6pta4HXgbmuO0zB3jGeX8RcIYyDW/mAC9rreu01rswhdpTO+fQBUEQBEEQBEEQehd2ArgsoMjlcbFzm8d9nG5dR4FUm68VBEEQBEEQBEEQbGAngFMetrk3EvW2j53XopS6USm1Sim1qqSkxMYhCYIgCIIgCIIg9D7sBHDFQI7L42xgv7d9lFIRQBJQZvO1aK0f1VpP1lpPTk9Pt3/0giAIgiAIgiAIvQildStB7MQdTEC2DTgD2AesBBZorTe57HMbMFZrfbNS6nLgYq3195RSo4EXMXVvA4CPgHytdVMbv68E2NOxPwuANKC0E94n1JBx8Y6MjWdkXLwjY+Mdu2MzUGstK3c26aQ5Us5b78jYeEbGxTsyNt6RsfFMh+dHn428tdaNSqmFwHtAOPCk1nqTUurXwCqt9WLgCeA5pVQhRnm73PnaTUqpV4FvgUbgtraCN+drOmUiV0qt0lpP7oz3CiVkXLwjY+MZGRfvyNh4R8ama+iMOVL+N96RsfGMjIt3ZGy8I2Pjmc4YF58BHIDWegmwxG3b/7jcrwUu9fLa+4D7OnCMgiAIgiAIgiAIAjYbeQuCIAiCIAiCIAiBJ5QDuEcDfQBBioyLd2RsPCPj4h0ZG+/I2AQv8r/xjoyNZ2RcvCNj4x0ZG890eFx8mpgIgiAIgiAIgiAIwUEoK3CCIAiCIAiCIAghRcgFcEqpc5RSW5VShUqpuwN9PIFEKfWkUuqwUuobl20pSqkPlFLbnbd9A3mMgUAplaOU+kQptVkptUkpdYdzu4yNUjFKqRVKqfXOsfmVc/tgpdRy59i8opSKCvSxBgKlVLhSaq1S6h3nYxkXQCm1Wym1USm1Tim1yrmt13+eghGZI1uQOdIzMkd6R+bItpE50jNdMUeGVACnlAoH/gmcC4wC5iulRgX2qALK08A5btvuBj7SWudj+vL1xgm8EfiJ1nokcBJwm/M8kbGBOmC21no8MAE4Ryl1EvBH4C/OsSkHrg/gMQaSO4DNLo9lXFo4XWs9wcUaWT5PQYbMka14GpkjPSFzpHdkjmwbmSO906lzZEgFcJiG4YVa651a63rgZWBOgI8pYGitP8X05XNlDvCM8/4zwEXdelBBgNb6gNZ6jfN+JebLJgsZG7ThuPNhpPNHA7OBRc7tvXJslFLZwHeBx52PFTIubdHrP09BiMyRLsgc6RmZI70jc6R3ZI70mw59nkItgMsCilweFzu3CS3001ofAPMlDWQE+HgCilJqEFAALEfGBmhOgVgHHAY+AHYAFVrrRucuvfVz9Vfgp4DD+TgVGRcLDbyvlFqtlLrRuU0+T8GHzJG+kfPWBZkjWyNzpFdkjvROp8+Rthp59yCUh21isyl4RCmVALwG/FBrfcwsFgla6yZgglIqGXgDGOlpt+49qsCilDofOKy1Xq2UmmVt9rBrrxoXF2ZorfcrpTKAD5RSWwJ9QIJH5JwVbCNzpGdkjmyNzJE+6fQ5MtQUuGIgx+VxNrA/QMcSrBxSSvUHcN4eDvDxBASlVCRmYnpBa/26c7OMjQta6wpgKaYGIlkpZS349MbP1QzgQqXUbkza2WzMamNvHxcAtNb7nbeHMRc0U5HPUzAic6Rv5LxF5kg7yBx5AjJHtkFXzJGhFsCtBPKdrjdRwOXA4gAfU7CxGLjGef8a4K0AHktAcOZlPwFs1lo/4PKUjI1S6c5VRZRSscCZmPqHT4BLnLv1urHRWt+jtc7WWg/CfK98rLW+gl4+LgBKqXilVB/rPnA28A3yeQpGZI70Ta8/b2WO9I7MkZ6ROdI7XTVHhlwjb6XUeZioPxx4Umt9X4APKWAopV4CZgFpwCHgf4E3gVeBXGAvcKnW2r2IO6RRSp0CfAZspCVX+78xOf69fWzGYYppwzELPK9qrX+tlBqCWVVLAdYCV2qt6wJ3pIHDmR5yp9b6fBkXcI7BG86HEcCLWuv7lFKp9PLPUzAic2QLMkd6RuZI78gc6RuZI0+kq+bIkAvgBEEQBEEQBEEQQpVQS6EUBEEQBEEQBEEIWSSAEwRBEARBEARB6CFIACcIgiAIgiAIgtBDkABOEARBEARBEAShhyABnCAIgiAIgiAIQg9BAjhBEARBEARBEIQeggRwgiAIgiAIgiAIPQQJ4ARBEARBEARBEHoI/x9krxjQwKMqNwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "show_history(history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "predict = model.evaluate_generator(validation_data, steps=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loss on test data set: 0.08317. Accuracy on test dataset: 95.70%\n"
     ]
    }
   ],
   "source": [
    "print(f\"Loss on test data set: {predict[0]:.5f}. Accuracy on test dataset: {predict[1] * 100:.2f}%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "def show_random_predictions(model, X_test):\n",
    "    predictions = model.predict(X_test)\n",
    "    print(predictions)\n",
    "    for i, row in enumerate(predictions):\n",
    "        is_parasitized = row[0] <= row[1]\n",
    "        label = \"Parasitized\" if is_parasitized else \"Uninfected\"\n",
    "        plt.title(label)\n",
    "        plt.imshow(X_test[i])\n",
    "        plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[4.0698182e-04 9.9959308e-01]\n",
      " [9.3226737e-01 6.7732587e-02]\n",
      " [3.1645693e-02 9.6835428e-01]\n",
      " [9.9305880e-01 6.9411620e-03]\n",
      " [9.7104204e-01 2.8957920e-02]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "random_data = X_test[np.random.randint(X_test.shape[0], size=5), :]\n",
    "show_random_predictions(model, random_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Training model using k-fold approach\n",
    "Kfold provides train/test indices to split data in train/test sets. Split dataset into k consecutive folds (without shuffling by default). Each fold is then used once as a validation while the k - 1 remaining folds form the training set.\n",
    "K-Fold is a popular and easy to understand, it generally results in a less biased model compare to other methods. Because it ensures that every observation from the original dataset has the chance of appearing in training and test set. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "split_num = 5\n",
    "kf = KFold(n_splits = 5, shuffle = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_categorical = np_utils.to_categorical(y, num_classes = 2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "def kfold_training(model, X: np.ndarray, y: np.ndarray, kfold_splitter: KFold) -> list:\n",
    "    scores = []\n",
    "    for _ in range(split_num):\n",
    "        kfold_indcies = next(kfold_splitter.split(X), None)\n",
    "        \n",
    "        X_train = X[kfold_indcies[0]]\n",
    "        y_train = y_categorical[kfold_indcies[0]]\n",
    "        X_test = X[kfold_indcies[1]]\n",
    "        y_test = y_categorical[kfold_indcies[1]]\n",
    "        \n",
    "        steps_per_epoch = X_train.shape[0] // 64\n",
    "\n",
    "        train_data = train_image_generator.flow(X_train, y_train, batch_size=64, shuffle=True)\n",
    "        validation_data = validation_image_generator.flow(X_test, y_test, batch_size=64, shuffle=True)\n",
    "\n",
    "        history = model.fit_generator(\n",
    "            train_data, \n",
    "            epochs=50,\n",
    "            steps_per_epoch=steps_per_epoch,\n",
    "            validation_data=validation_data,\n",
    "            validation_steps=6,\n",
    "            verbose=1\n",
    "        )\n",
    "\n",
    "        predict = model.evaluate_generator(validation_data, steps=4)\n",
    "        scores.append(predict)\n",
    "    \n",
    "    return scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1499 - accuracy: 0.9461 - val_loss: 0.1513 - val_accuracy: 0.9661\n",
      "Epoch 2/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1516 - accuracy: 0.9454 - val_loss: 0.0324 - val_accuracy: 0.9740\n",
      "Epoch 3/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1548 - accuracy: 0.9438 - val_loss: 0.0652 - val_accuracy: 0.9714\n",
      "Epoch 4/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1429 - accuracy: 0.9476 - val_loss: 0.1045 - val_accuracy: 0.9766\n",
      "Epoch 5/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1434 - accuracy: 0.9493 - val_loss: 0.1398 - val_accuracy: 0.9479\n",
      "Epoch 6/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1492 - accuracy: 0.9481 - val_loss: 0.2010 - val_accuracy: 0.9635\n",
      "Epoch 7/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1466 - accuracy: 0.9473 - val_loss: 0.0390 - val_accuracy: 0.9792\n",
      "Epoch 8/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1460 - accuracy: 0.9467 - val_loss: 0.0264 - val_accuracy: 0.9714\n",
      "Epoch 9/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1507 - accuracy: 0.9448 - val_loss: 0.0272 - val_accuracy: 0.9766\n",
      "Epoch 10/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1476 - accuracy: 0.9470 - val_loss: 0.0826 - val_accuracy: 0.9661\n",
      "Epoch 11/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1476 - accuracy: 0.9470 - val_loss: 0.0230 - val_accuracy: 0.9818\n",
      "Epoch 12/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1476 - accuracy: 0.9486 - val_loss: 0.1272 - val_accuracy: 0.9714\n",
      "Epoch 13/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1452 - accuracy: 0.9484 - val_loss: 0.0569 - val_accuracy: 0.9635\n",
      "Epoch 14/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1448 - accuracy: 0.9467 - val_loss: 0.1668 - val_accuracy: 0.9635\n",
      "Epoch 15/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1513 - accuracy: 0.9467 - val_loss: 0.0997 - val_accuracy: 0.9787\n",
      "Epoch 16/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1475 - accuracy: 0.9454 - val_loss: 0.1278 - val_accuracy: 0.9531\n",
      "Epoch 17/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1529 - accuracy: 0.9450 - val_loss: 0.1281 - val_accuracy: 0.9818\n",
      "Epoch 18/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1489 - accuracy: 0.9452 - val_loss: 0.0834 - val_accuracy: 0.9844\n",
      "Epoch 19/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1467 - accuracy: 0.9475 - val_loss: 0.1576 - val_accuracy: 0.9661\n",
      "Epoch 20/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1443 - accuracy: 0.9472 - val_loss: 0.2240 - val_accuracy: 0.9661\n",
      "Epoch 21/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1421 - accuracy: 0.9499 - val_loss: 0.1297 - val_accuracy: 0.9661\n",
      "Epoch 22/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1450 - accuracy: 0.9496 - val_loss: 0.0641 - val_accuracy: 0.9609\n",
      "Epoch 23/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1475 - accuracy: 0.9476 - val_loss: 0.0688 - val_accuracy: 0.9661\n",
      "Epoch 24/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1439 - accuracy: 0.9473 - val_loss: 0.0600 - val_accuracy: 0.9661\n",
      "Epoch 25/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1477 - accuracy: 0.9470 - val_loss: 0.1077 - val_accuracy: 0.9557\n",
      "Epoch 26/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1434 - accuracy: 0.9476 - val_loss: 0.0124 - val_accuracy: 0.9844\n",
      "Epoch 27/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1504 - accuracy: 0.9460 - val_loss: 0.1685 - val_accuracy: 0.9688\n",
      "Epoch 28/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1444 - accuracy: 0.9483 - val_loss: 0.0269 - val_accuracy: 0.9740\n",
      "Epoch 29/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1485 - accuracy: 0.9470 - val_loss: 0.0217 - val_accuracy: 0.9604\n",
      "Epoch 30/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1457 - accuracy: 0.9470 - val_loss: 0.1469 - val_accuracy: 0.9688\n",
      "Epoch 31/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1478 - accuracy: 0.9476 - val_loss: 0.0537 - val_accuracy: 0.9688\n",
      "Epoch 32/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1457 - accuracy: 0.9487 - val_loss: 0.0895 - val_accuracy: 0.9792\n",
      "Epoch 33/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1383 - accuracy: 0.9513 - val_loss: 0.0447 - val_accuracy: 0.9766\n",
      "Epoch 34/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1432 - accuracy: 0.9491 - val_loss: 0.1148 - val_accuracy: 0.9505\n",
      "Epoch 35/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1406 - accuracy: 0.9488 - val_loss: 0.0500 - val_accuracy: 0.9531\n",
      "Epoch 36/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1397 - accuracy: 0.9498 - val_loss: 0.0492 - val_accuracy: 0.9661\n",
      "Epoch 37/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1441 - accuracy: 0.9475 - val_loss: 0.0806 - val_accuracy: 0.9688\n",
      "Epoch 38/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1483 - accuracy: 0.9479 - val_loss: 0.2145 - val_accuracy: 0.9609\n",
      "Epoch 39/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1447 - accuracy: 0.9483 - val_loss: 0.0696 - val_accuracy: 0.9583\n",
      "Epoch 40/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1462 - accuracy: 0.9483 - val_loss: 0.0366 - val_accuracy: 0.9661\n",
      "Epoch 41/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1436 - accuracy: 0.9485 - val_loss: 0.1369 - val_accuracy: 0.9714\n",
      "Epoch 42/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1450 - accuracy: 0.9469 - val_loss: 0.0374 - val_accuracy: 0.9688\n",
      "Epoch 43/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1454 - accuracy: 0.9480 - val_loss: 0.0449 - val_accuracy: 0.9792\n",
      "Epoch 44/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1379 - accuracy: 0.9509 - val_loss: 0.1228 - val_accuracy: 0.9512\n",
      "Epoch 45/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1453 - accuracy: 0.9475 - val_loss: 0.0800 - val_accuracy: 0.9740\n",
      "Epoch 46/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1474 - accuracy: 0.9486 - val_loss: 0.1538 - val_accuracy: 0.9505\n",
      "Epoch 47/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1436 - accuracy: 0.9468 - val_loss: 0.1512 - val_accuracy: 0.9766\n",
      "Epoch 48/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1429 - accuracy: 0.9484 - val_loss: 0.1001 - val_accuracy: 0.9688\n",
      "Epoch 49/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1427 - accuracy: 0.9489 - val_loss: 0.0283 - val_accuracy: 0.9688\n",
      "Epoch 50/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1431 - accuracy: 0.9481 - val_loss: 0.1494 - val_accuracy: 0.9635\n",
      "Epoch 1/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1450 - accuracy: 0.9479 - val_loss: 0.1056 - val_accuracy: 0.9609\n",
      "Epoch 2/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1405 - accuracy: 0.9501 - val_loss: 0.1423 - val_accuracy: 0.9531\n",
      "Epoch 3/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1421 - accuracy: 0.9493 - val_loss: 0.0420 - val_accuracy: 0.9740\n",
      "Epoch 4/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1413 - accuracy: 0.9491 - val_loss: 0.1111 - val_accuracy: 0.9792\n",
      "Epoch 5/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1433 - accuracy: 0.9485 - val_loss: 0.1438 - val_accuracy: 0.9740\n",
      "Epoch 6/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1422 - accuracy: 0.9498 - val_loss: 0.0957 - val_accuracy: 0.9505\n",
      "Epoch 7/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1413 - accuracy: 0.9494 - val_loss: 0.0343 - val_accuracy: 0.9635\n",
      "Epoch 8/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1367 - accuracy: 0.9504 - val_loss: 0.1453 - val_accuracy: 0.9505\n",
      "Epoch 9/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1398 - accuracy: 0.9494 - val_loss: 0.0702 - val_accuracy: 0.9583\n",
      "Epoch 10/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1402 - accuracy: 0.9485 - val_loss: 0.0369 - val_accuracy: 0.9714\n",
      "Epoch 11/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1411 - accuracy: 0.9495 - val_loss: 0.1507 - val_accuracy: 0.9766\n",
      "Epoch 12/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1433 - accuracy: 0.9487 - val_loss: 0.0995 - val_accuracy: 0.9792\n",
      "Epoch 13/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1464 - accuracy: 0.9486 - val_loss: 0.1059 - val_accuracy: 0.9557\n",
      "Epoch 14/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1382 - accuracy: 0.9505 - val_loss: 0.0365 - val_accuracy: 0.9766\n",
      "Epoch 15/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1400 - accuracy: 0.9498 - val_loss: 0.1319 - val_accuracy: 0.9756\n",
      "Epoch 16/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1372 - accuracy: 0.9498 - val_loss: 0.1869 - val_accuracy: 0.9635\n",
      "Epoch 17/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1347 - accuracy: 0.9509 - val_loss: 0.0198 - val_accuracy: 0.9531\n",
      "Epoch 18/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1414 - accuracy: 0.9506 - val_loss: 0.1361 - val_accuracy: 0.9714\n",
      "Epoch 19/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1394 - accuracy: 0.9496 - val_loss: 0.0942 - val_accuracy: 0.9661\n",
      "Epoch 20/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1458 - accuracy: 0.9481 - val_loss: 0.1967 - val_accuracy: 0.9583\n",
      "Epoch 21/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1376 - accuracy: 0.9515 - val_loss: 0.1250 - val_accuracy: 0.9714\n",
      "Epoch 22/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1383 - accuracy: 0.9506 - val_loss: 0.0532 - val_accuracy: 0.9609\n",
      "Epoch 23/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1409 - accuracy: 0.9505 - val_loss: 0.0402 - val_accuracy: 0.9714\n",
      "Epoch 24/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1374 - accuracy: 0.9499 - val_loss: 0.1118 - val_accuracy: 0.9583\n",
      "Epoch 25/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1350 - accuracy: 0.9519 - val_loss: 0.1184 - val_accuracy: 0.9661\n",
      "Epoch 26/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1424 - accuracy: 0.9482 - val_loss: 0.0470 - val_accuracy: 0.9792\n",
      "Epoch 27/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1372 - accuracy: 0.9518 - val_loss: 0.1129 - val_accuracy: 0.9609\n",
      "Epoch 28/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1399 - accuracy: 0.9496 - val_loss: 0.1104 - val_accuracy: 0.9740\n",
      "Epoch 29/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1345 - accuracy: 0.9520 - val_loss: 0.0132 - val_accuracy: 0.9665\n",
      "Epoch 30/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1343 - accuracy: 0.9520 - val_loss: 0.0858 - val_accuracy: 0.9531\n",
      "Epoch 31/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1358 - accuracy: 0.9514 - val_loss: 0.2332 - val_accuracy: 0.9557\n",
      "Epoch 32/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1400 - accuracy: 0.9497 - val_loss: 0.0669 - val_accuracy: 0.9583\n",
      "Epoch 33/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1345 - accuracy: 0.9517 - val_loss: 0.1109 - val_accuracy: 0.9661\n",
      "Epoch 34/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1396 - accuracy: 0.9518 - val_loss: 0.0607 - val_accuracy: 0.9740\n",
      "Epoch 35/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1408 - accuracy: 0.9496 - val_loss: 0.0552 - val_accuracy: 0.9609\n",
      "Epoch 36/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1438 - accuracy: 0.9487 - val_loss: 0.0229 - val_accuracy: 0.9688\n",
      "Epoch 37/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1371 - accuracy: 0.9509 - val_loss: 0.0397 - val_accuracy: 0.9609\n",
      "Epoch 38/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1376 - accuracy: 0.9516 - val_loss: 0.0574 - val_accuracy: 0.9766\n",
      "Epoch 39/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1384 - accuracy: 0.9515 - val_loss: 0.0967 - val_accuracy: 0.9688\n",
      "Epoch 40/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1392 - accuracy: 0.9492 - val_loss: 0.0669 - val_accuracy: 0.9740\n",
      "Epoch 41/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1348 - accuracy: 0.9527 - val_loss: 0.0377 - val_accuracy: 0.9818\n",
      "Epoch 42/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1410 - accuracy: 0.9494 - val_loss: 0.1885 - val_accuracy: 0.9427\n",
      "Epoch 43/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1386 - accuracy: 0.9498 - val_loss: 0.0301 - val_accuracy: 0.9740\n",
      "Epoch 44/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1351 - accuracy: 0.9514 - val_loss: 0.0397 - val_accuracy: 0.9695\n",
      "Epoch 45/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1379 - accuracy: 0.9506 - val_loss: 0.0897 - val_accuracy: 0.9531\n",
      "Epoch 46/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1361 - accuracy: 0.9501 - val_loss: 0.1220 - val_accuracy: 0.9766\n",
      "Epoch 47/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1382 - accuracy: 0.9497 - val_loss: 0.1639 - val_accuracy: 0.9557\n",
      "Epoch 48/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1407 - accuracy: 0.9499 - val_loss: 0.1136 - val_accuracy: 0.9688\n",
      "Epoch 49/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1353 - accuracy: 0.9520 - val_loss: 0.1018 - val_accuracy: 0.9740\n",
      "Epoch 50/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1341 - accuracy: 0.9526 - val_loss: 0.3007 - val_accuracy: 0.9505\n",
      "Epoch 1/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1409 - accuracy: 0.9497 - val_loss: 0.0282 - val_accuracy: 0.9714\n",
      "Epoch 2/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1383 - accuracy: 0.9505 - val_loss: 0.0328 - val_accuracy: 0.9688\n",
      "Epoch 3/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1397 - accuracy: 0.9502 - val_loss: 0.1346 - val_accuracy: 0.9844\n",
      "Epoch 4/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1395 - accuracy: 0.9491 - val_loss: 0.1304 - val_accuracy: 0.9740\n",
      "Epoch 5/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1464 - accuracy: 0.9473 - val_loss: 0.1000 - val_accuracy: 0.9635\n",
      "Epoch 6/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1411 - accuracy: 0.9493 - val_loss: 0.0421 - val_accuracy: 0.9714\n",
      "Epoch 7/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1460 - accuracy: 0.9479 - val_loss: 0.0639 - val_accuracy: 0.9792\n",
      "Epoch 8/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1430 - accuracy: 0.9477 - val_loss: 0.0280 - val_accuracy: 0.9792\n",
      "Epoch 9/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1433 - accuracy: 0.9486 - val_loss: 0.0787 - val_accuracy: 0.9792\n",
      "Epoch 10/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1388 - accuracy: 0.9505 - val_loss: 0.0756 - val_accuracy: 0.9661\n",
      "Epoch 11/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1379 - accuracy: 0.9511 - val_loss: 0.1354 - val_accuracy: 0.9401\n",
      "Epoch 12/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1372 - accuracy: 0.9495 - val_loss: 0.0876 - val_accuracy: 0.9688\n",
      "Epoch 13/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1335 - accuracy: 0.9525 - val_loss: 0.0414 - val_accuracy: 0.9609\n",
      "Epoch 14/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1435 - accuracy: 0.9484 - val_loss: 0.0803 - val_accuracy: 0.9688\n",
      "Epoch 15/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1398 - accuracy: 0.9498 - val_loss: 0.1306 - val_accuracy: 0.9726\n",
      "Epoch 16/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1402 - accuracy: 0.9493 - val_loss: 0.0481 - val_accuracy: 0.9766\n",
      "Epoch 17/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1375 - accuracy: 0.9506 - val_loss: 0.0609 - val_accuracy: 0.9792\n",
      "Epoch 18/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1364 - accuracy: 0.9506 - val_loss: 0.0337 - val_accuracy: 0.9688\n",
      "Epoch 19/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1331 - accuracy: 0.9505 - val_loss: 0.0544 - val_accuracy: 0.9609\n",
      "Epoch 20/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1440 - accuracy: 0.9489 - val_loss: 0.0862 - val_accuracy: 0.9583\n",
      "Epoch 21/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1419 - accuracy: 0.9491 - val_loss: 0.0990 - val_accuracy: 0.9714\n",
      "Epoch 22/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1395 - accuracy: 0.9489 - val_loss: 0.0191 - val_accuracy: 0.9661\n",
      "Epoch 23/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1441 - accuracy: 0.9482 - val_loss: 0.0336 - val_accuracy: 0.9714\n",
      "Epoch 24/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1372 - accuracy: 0.9497 - val_loss: 0.1907 - val_accuracy: 0.9609\n",
      "Epoch 25/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1382 - accuracy: 0.9509 - val_loss: 0.1042 - val_accuracy: 0.9453\n",
      "Epoch 26/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1411 - accuracy: 0.9495 - val_loss: 0.0523 - val_accuracy: 0.9818\n",
      "Epoch 27/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1392 - accuracy: 0.9501 - val_loss: 0.0759 - val_accuracy: 0.9688\n",
      "Epoch 28/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1382 - accuracy: 0.9491 - val_loss: 0.0494 - val_accuracy: 0.9609\n",
      "Epoch 29/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1369 - accuracy: 0.9523 - val_loss: 0.0256 - val_accuracy: 0.9756\n",
      "Epoch 30/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1392 - accuracy: 0.9506 - val_loss: 0.1403 - val_accuracy: 0.9714\n",
      "Epoch 31/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1378 - accuracy: 0.9512 - val_loss: 0.0960 - val_accuracy: 0.9661\n",
      "Epoch 32/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1415 - accuracy: 0.9490 - val_loss: 0.1085 - val_accuracy: 0.9688\n",
      "Epoch 33/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1383 - accuracy: 0.9506 - val_loss: 0.0498 - val_accuracy: 0.9844\n",
      "Epoch 34/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1388 - accuracy: 0.9512 - val_loss: 0.0889 - val_accuracy: 0.9740\n",
      "Epoch 35/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1371 - accuracy: 0.9510 - val_loss: 0.0553 - val_accuracy: 0.9505\n",
      "Epoch 36/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1348 - accuracy: 0.9515 - val_loss: 0.0692 - val_accuracy: 0.9766\n",
      "Epoch 37/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1505 - accuracy: 0.9467 - val_loss: 0.2621 - val_accuracy: 0.9661\n",
      "Epoch 38/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1391 - accuracy: 0.9489 - val_loss: 0.0514 - val_accuracy: 0.9766\n",
      "Epoch 39/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1385 - accuracy: 0.9510 - val_loss: 0.0150 - val_accuracy: 0.9766\n",
      "Epoch 40/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1402 - accuracy: 0.9500 - val_loss: 0.0597 - val_accuracy: 0.9583\n",
      "Epoch 41/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1442 - accuracy: 0.9496 - val_loss: 0.0900 - val_accuracy: 0.9688\n",
      "Epoch 42/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1386 - accuracy: 0.9512 - val_loss: 0.2949 - val_accuracy: 0.9583\n",
      "Epoch 43/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1310 - accuracy: 0.9530 - val_loss: 0.0600 - val_accuracy: 0.9818\n",
      "Epoch 44/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1396 - accuracy: 0.9482 - val_loss: 0.0325 - val_accuracy: 0.9787\n",
      "Epoch 45/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1361 - accuracy: 0.9520 - val_loss: 0.0995 - val_accuracy: 0.9818\n",
      "Epoch 46/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1402 - accuracy: 0.9501 - val_loss: 0.0596 - val_accuracy: 0.9818\n",
      "Epoch 47/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1335 - accuracy: 0.9527 - val_loss: 0.1401 - val_accuracy: 0.9661\n",
      "Epoch 48/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1345 - accuracy: 0.9521 - val_loss: 0.2501 - val_accuracy: 0.9609\n",
      "Epoch 49/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1372 - accuracy: 0.9516 - val_loss: 0.0638 - val_accuracy: 0.9688\n",
      "Epoch 50/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1375 - accuracy: 0.9514 - val_loss: 0.1308 - val_accuracy: 0.9609\n",
      "Epoch 1/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1399 - accuracy: 0.9506 - val_loss: 0.1528 - val_accuracy: 0.9688\n",
      "Epoch 2/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1358 - accuracy: 0.9519 - val_loss: 0.1476 - val_accuracy: 0.9661\n",
      "Epoch 3/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1360 - accuracy: 0.9514 - val_loss: 0.0868 - val_accuracy: 0.9661\n",
      "Epoch 4/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1377 - accuracy: 0.9517 - val_loss: 0.1097 - val_accuracy: 0.9792\n",
      "Epoch 5/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1371 - accuracy: 0.9514 - val_loss: 0.1730 - val_accuracy: 0.9557\n",
      "Epoch 6/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1405 - accuracy: 0.9501 - val_loss: 0.0730 - val_accuracy: 0.9766\n",
      "Epoch 7/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1304 - accuracy: 0.9532 - val_loss: 0.0638 - val_accuracy: 0.9661\n",
      "Epoch 8/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1323 - accuracy: 0.9516 - val_loss: 0.0732 - val_accuracy: 0.9688\n",
      "Epoch 9/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1312 - accuracy: 0.9520 - val_loss: 0.0847 - val_accuracy: 0.9661\n",
      "Epoch 10/50\n",
      "344/344 [==============================] - ETA: 0s - loss: 0.1344 - accuracy: 0.95 - 13s 38ms/step - loss: 0.1345 - accuracy: 0.9523 - val_loss: 0.2094 - val_accuracy: 0.9609\n",
      "Epoch 11/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1351 - accuracy: 0.9516 - val_loss: 0.0815 - val_accuracy: 0.9661\n",
      "Epoch 12/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1360 - accuracy: 0.9520 - val_loss: 0.0679 - val_accuracy: 0.9688\n",
      "Epoch 13/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1324 - accuracy: 0.9520 - val_loss: 0.0829 - val_accuracy: 0.9714\n",
      "Epoch 14/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1321 - accuracy: 0.9515 - val_loss: 0.0565 - val_accuracy: 0.9583\n",
      "Epoch 15/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1299 - accuracy: 0.9524 - val_loss: 0.0646 - val_accuracy: 0.9756\n",
      "Epoch 16/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1291 - accuracy: 0.9546 - val_loss: 0.0866 - val_accuracy: 0.9557\n",
      "Epoch 17/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1382 - accuracy: 0.9496 - val_loss: 0.0456 - val_accuracy: 0.9661\n",
      "Epoch 18/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1325 - accuracy: 0.9521 - val_loss: 0.0699 - val_accuracy: 0.9583\n",
      "Epoch 19/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1383 - accuracy: 0.9495 - val_loss: 0.1215 - val_accuracy: 0.9609\n",
      "Epoch 20/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1387 - accuracy: 0.9504 - val_loss: 0.0913 - val_accuracy: 0.9609\n",
      "Epoch 21/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1361 - accuracy: 0.9517 - val_loss: 0.0513 - val_accuracy: 0.9661\n",
      "Epoch 22/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1377 - accuracy: 0.9514 - val_loss: 0.1197 - val_accuracy: 0.9609\n",
      "Epoch 23/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1393 - accuracy: 0.9503 - val_loss: 0.1369 - val_accuracy: 0.9635\n",
      "Epoch 24/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1386 - accuracy: 0.9508 - val_loss: 0.0698 - val_accuracy: 0.9688\n",
      "Epoch 25/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1315 - accuracy: 0.9510 - val_loss: 0.0234 - val_accuracy: 0.9740\n",
      "Epoch 26/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1307 - accuracy: 0.9539 - val_loss: 0.0473 - val_accuracy: 0.9740\n",
      "Epoch 27/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1360 - accuracy: 0.9513 - val_loss: 0.1127 - val_accuracy: 0.9688\n",
      "Epoch 28/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1387 - accuracy: 0.9513 - val_loss: 0.0465 - val_accuracy: 0.9714\n",
      "Epoch 29/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1355 - accuracy: 0.9528 - val_loss: 0.0493 - val_accuracy: 0.9756\n",
      "Epoch 30/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1364 - accuracy: 0.9512 - val_loss: 0.1352 - val_accuracy: 0.9583\n",
      "Epoch 31/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1313 - accuracy: 0.9533 - val_loss: 0.0746 - val_accuracy: 0.9688\n",
      "Epoch 32/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1291 - accuracy: 0.9546 - val_loss: 0.0963 - val_accuracy: 0.9479\n",
      "Epoch 33/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1346 - accuracy: 0.9509 - val_loss: 0.0493 - val_accuracy: 0.9792\n",
      "Epoch 34/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1330 - accuracy: 0.9527 - val_loss: 0.1631 - val_accuracy: 0.9505\n",
      "Epoch 35/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1361 - accuracy: 0.9520 - val_loss: 0.1050 - val_accuracy: 0.9505\n",
      "Epoch 36/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1296 - accuracy: 0.9522 - val_loss: 0.0670 - val_accuracy: 0.9740\n",
      "Epoch 37/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1344 - accuracy: 0.9508 - val_loss: 0.0441 - val_accuracy: 0.9714\n",
      "Epoch 38/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1307 - accuracy: 0.9536 - val_loss: 0.0391 - val_accuracy: 0.9505\n",
      "Epoch 39/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1379 - accuracy: 0.9507 - val_loss: 0.1458 - val_accuracy: 0.9505\n",
      "Epoch 40/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1436 - accuracy: 0.9487 - val_loss: 0.0456 - val_accuracy: 0.9583\n",
      "Epoch 41/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1324 - accuracy: 0.9534 - val_loss: 0.0342 - val_accuracy: 0.9635\n",
      "Epoch 42/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1269 - accuracy: 0.9546 - val_loss: 0.0403 - val_accuracy: 0.9714\n",
      "Epoch 43/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1286 - accuracy: 0.9537 - val_loss: 0.0261 - val_accuracy: 0.9844\n",
      "Epoch 44/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1295 - accuracy: 0.9536 - val_loss: 0.1811 - val_accuracy: 0.9604\n",
      "Epoch 45/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1331 - accuracy: 0.9509 - val_loss: 0.0762 - val_accuracy: 0.9609\n",
      "Epoch 46/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1314 - accuracy: 0.9533 - val_loss: 0.0252 - val_accuracy: 0.9766\n",
      "Epoch 47/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1340 - accuracy: 0.9524 - val_loss: 0.1788 - val_accuracy: 0.9635\n",
      "Epoch 48/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1342 - accuracy: 0.9509 - val_loss: 0.0171 - val_accuracy: 0.9714\n",
      "Epoch 49/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1336 - accuracy: 0.9522 - val_loss: 0.0219 - val_accuracy: 0.9766\n",
      "Epoch 50/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1355 - accuracy: 0.9511 - val_loss: 0.0425 - val_accuracy: 0.9453\n",
      "Epoch 1/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1388 - accuracy: 0.9508 - val_loss: 0.0259 - val_accuracy: 0.9844\n",
      "Epoch 2/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1372 - accuracy: 0.9509 - val_loss: 0.0390 - val_accuracy: 0.9870\n",
      "Epoch 3/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1359 - accuracy: 0.9519 - val_loss: 0.0271 - val_accuracy: 0.9661\n",
      "Epoch 4/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1375 - accuracy: 0.9494 - val_loss: 0.0806 - val_accuracy: 0.9688\n",
      "Epoch 5/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1311 - accuracy: 0.9514 - val_loss: 0.1829 - val_accuracy: 0.9557\n",
      "Epoch 6/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1380 - accuracy: 0.9495 - val_loss: 0.0737 - val_accuracy: 0.9609\n",
      "Epoch 7/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1313 - accuracy: 0.9535 - val_loss: 0.0529 - val_accuracy: 0.9740\n",
      "Epoch 8/50\n",
      "344/344 [==============================] - 13s 38ms/step - loss: 0.1396 - accuracy: 0.9493 - val_loss: 0.0479 - val_accuracy: 0.9661\n",
      "Epoch 9/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1379 - accuracy: 0.9497 - val_loss: 0.2541 - val_accuracy: 0.9688\n",
      "Epoch 10/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1301 - accuracy: 0.9532 - val_loss: 0.0704 - val_accuracy: 0.9818\n",
      "Epoch 11/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1387 - accuracy: 0.9496 - val_loss: 0.1637 - val_accuracy: 0.9714\n",
      "Epoch 12/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1433 - accuracy: 0.9493 - val_loss: 0.1624 - val_accuracy: 0.9531\n",
      "Epoch 13/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1377 - accuracy: 0.9507 - val_loss: 0.0345 - val_accuracy: 0.9844\n",
      "Epoch 14/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1355 - accuracy: 0.9501 - val_loss: 0.0666 - val_accuracy: 0.9688\n",
      "Epoch 15/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1298 - accuracy: 0.9540 - val_loss: 0.0741 - val_accuracy: 0.9512\n",
      "Epoch 16/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1380 - accuracy: 0.9517 - val_loss: 0.0630 - val_accuracy: 0.9531\n",
      "Epoch 17/50\n",
      "344/344 [==============================] - 12s 36ms/step - loss: 0.1327 - accuracy: 0.9533 - val_loss: 0.0641 - val_accuracy: 0.9792\n",
      "Epoch 18/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1367 - accuracy: 0.9515 - val_loss: 0.0775 - val_accuracy: 0.9635\n",
      "Epoch 19/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1304 - accuracy: 0.9526 - val_loss: 0.1423 - val_accuracy: 0.9609\n",
      "Epoch 20/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1331 - accuracy: 0.9507 - val_loss: 0.0990 - val_accuracy: 0.9766\n",
      "Epoch 21/50\n",
      "344/344 [==============================] - 13s 36ms/step - loss: 0.1355 - accuracy: 0.9512 - val_loss: 0.0176 - val_accuracy: 0.9557\n",
      "Epoch 22/50\n",
      "344/344 [==============================] - ETA: 0s - loss: 0.1383 - accuracy: 0.95 - 13s 37ms/step - loss: 0.1386 - accuracy: 0.9517 - val_loss: 0.0850 - val_accuracy: 0.9688\n",
      "Epoch 23/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1293 - accuracy: 0.9526 - val_loss: 0.1000 - val_accuracy: 0.9635\n",
      "Epoch 24/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1286 - accuracy: 0.9521 - val_loss: 0.1532 - val_accuracy: 0.9818\n",
      "Epoch 25/50\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1404 - accuracy: 0.9496 - val_loss: 0.1350 - val_accuracy: 0.9688\n",
      "Epoch 26/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1338 - accuracy: 0.9526 - val_loss: 0.0674 - val_accuracy: 0.9844\n",
      "Epoch 27/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1304 - accuracy: 0.9525 - val_loss: 0.0772 - val_accuracy: 0.9688\n",
      "Epoch 28/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1329 - accuracy: 0.9529 - val_loss: 0.1267 - val_accuracy: 0.9740\n",
      "Epoch 29/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1298 - accuracy: 0.9521 - val_loss: 0.0830 - val_accuracy: 0.9665\n",
      "Epoch 30/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1361 - accuracy: 0.9508 - val_loss: 0.0971 - val_accuracy: 0.9661\n",
      "Epoch 31/50\n",
      "344/344 [==============================] - 14s 39ms/step - loss: 0.1284 - accuracy: 0.9537 - val_loss: 0.0448 - val_accuracy: 0.9818\n",
      "Epoch 32/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1313 - accuracy: 0.9516 - val_loss: 0.0191 - val_accuracy: 0.9714\n",
      "Epoch 33/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1361 - accuracy: 0.9509 - val_loss: 0.0152 - val_accuracy: 0.9740\n",
      "Epoch 34/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1308 - accuracy: 0.9516 - val_loss: 0.1435 - val_accuracy: 0.9635\n",
      "Epoch 35/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1293 - accuracy: 0.9534 - val_loss: 0.1362 - val_accuracy: 0.9818\n",
      "Epoch 36/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1303 - accuracy: 0.9538 - val_loss: 0.1702 - val_accuracy: 0.9557\n",
      "Epoch 37/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1353 - accuracy: 0.9529 - val_loss: 0.0216 - val_accuracy: 0.9766\n",
      "Epoch 38/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1321 - accuracy: 0.9528 - val_loss: 0.0595 - val_accuracy: 0.9635\n",
      "Epoch 39/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1351 - accuracy: 0.9512 - val_loss: 0.0656 - val_accuracy: 0.9792\n",
      "Epoch 40/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1359 - accuracy: 0.9519 - val_loss: 0.0694 - val_accuracy: 0.9609\n",
      "Epoch 41/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1344 - accuracy: 0.9515 - val_loss: 0.1968 - val_accuracy: 0.9479\n",
      "Epoch 42/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1354 - accuracy: 0.9519 - val_loss: 0.2060 - val_accuracy: 0.9635\n",
      "Epoch 43/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1299 - accuracy: 0.9546 - val_loss: 0.0453 - val_accuracy: 0.9792\n",
      "Epoch 44/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1311 - accuracy: 0.9530 - val_loss: 0.0716 - val_accuracy: 0.9695\n",
      "Epoch 45/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1329 - accuracy: 0.9523 - val_loss: 0.0825 - val_accuracy: 0.9661\n",
      "Epoch 46/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1293 - accuracy: 0.9519 - val_loss: 0.0831 - val_accuracy: 0.9661\n",
      "Epoch 47/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1339 - accuracy: 0.9518 - val_loss: 0.0617 - val_accuracy: 0.9818\n",
      "Epoch 48/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1396 - accuracy: 0.9501 - val_loss: 0.1202 - val_accuracy: 0.9557\n",
      "Epoch 49/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1305 - accuracy: 0.9532 - val_loss: 0.0256 - val_accuracy: 0.9714\n",
      "Epoch 50/50\n",
      "344/344 [==============================] - 13s 37ms/step - loss: 0.1301 - accuracy: 0.9536 - val_loss: 0.0337 - val_accuracy: 0.9766\n"
     ]
    }
   ],
   "source": [
    "scores = kfold_training(model, X, y_categorical, kf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scores from each Iteration (loss/accuracy)\n",
      "Results from 1 iteration. Loss: 0.029980065301060677, Accuracy: 0.95703125\n",
      "Results from 2 iteration. Loss: 0.11633279919624329, Accuracy: 0.96484375\n",
      "Results from 3 iteration. Loss: 0.28844285011291504, Accuracy: 0.93359375\n",
      "Results from 4 iteration. Loss: 0.09662795811891556, Accuracy: 0.96484375\n",
      "Results from 5 iteration. Loss: 0.120735302567482, Accuracy: 0.97265625\n"
     ]
    }
   ],
   "source": [
    "print('Scores from each Iteration (loss/accuracy)')\n",
    "for i, score in enumerate(scores):\n",
    "    print(f\"Results from {i+1} iteration. Loss: {score[0]}, Accuracy: {score[1]}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Average K-Fold Score : 0.95859375\n"
     ]
    }
   ],
   "source": [
    "print('Average K-Fold Score :' , np.mean([x[1] for x in scores]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comparison with classical methods"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As the raise of CNNs dominated image classification there are planty of classical methods that can be used for that kind of problems. In the below section you can find some classical methods and how they perform on Malaria cells dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "from sklearn import svm, metrics\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import MinMaxScaler, StandardScaler"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Data preprocessing\n",
    "Preprocessing for SVM is a bit more complex than preprocessing for CNNs. Besides resizing, images have to be converted to grayscale and normalized. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [],
   "source": [
    "def preprocess_images_svm(images: list, base_path: str, img_label: int, resize: tuple = (50, 50)) -> tuple:\n",
    "    images = [f for f in images if f.endswith('.png')]  # filter out only png images\n",
    "    X, y = [], []\n",
    "    \n",
    "    for image in images:\n",
    "        img_arr = plt.imread(os.path.join(base_path, image))\n",
    "        \n",
    "        resized_img = cv2.resize(img_arr, resize)  # rescale\n",
    "        gray_img = cv2.cvtColor(resized_img, cv2.COLOR_BGR2GRAY)  # convert to grayscale\n",
    "        \n",
    "        # range starts from 0.1 to eliminate all black background\n",
    "        counts, bins = np.histogram(gray_img, bins=10, range=(0.1, 1))\n",
    "        histnorm_img = counts / np.sum(counts)  # histnorm images\n",
    "        \n",
    "        X.append(histnorm_img)\n",
    "        y.append(img_label)\n",
    "\n",
    "    return X, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_p_svm, y_p_svm = preprocess_images_svm(parasitized_cells, parasitized_dir, 1)\n",
    "X_u_svm, y_u_svm = preprocess_images_svm(uninfected_cells, uninfected_dir, 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Combine both preprocessed parasitized and uninfected images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_svm = np.array([*X_p_svm, *X_u_svm])\n",
    "y_svm = np.array([*y_p_svm, *y_u_svm])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_svm, X_test_svm, y_train_svm, y_test_svm = train_test_split(X_svm, y_svm, test_size=0.3, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Transform features by scaling each feature to a (0, 1) range. It is prerequisite for using PCA.\n",
    "This estimator scales and translates each feature individually such that it is in the given range on the training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler()\n",
    "X_train_stand = scaler.fit_transform(X_train_svm)\n",
    "X_test_stand = scaler.fit_transform(X_test_svm)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "SVMs are not working good with complicated, multi-feature data. By using PCA we are projecting data to a lower dimensional space, so SVM will perform a lot better.\n",
    "As `n_components` let's set the variance of the input that is supposed to be explained by the generated components.\n",
    "Typically, we want the explained variance to be between 95–99%. After some tests it apprears that `0.99` is the best fit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=0.99)\n",
    "pca.fit(X_train_svm_stand)\n",
    "pca_X_train = pca.transform(X_train_svm_stand)\n",
    "pca_X_test = pca.transform(X_test_svm_stand)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "SVM method was tested on different sets of paramaters, gamma = (0.01, 0.3), C = (10, 2000). The best set of parameteres was used in notebook. \n",
    "Used kernel: `rbf`. SVM wasn't performing well with linear and poly kernels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "SVC(C=2000, break_ties=False, cache_size=200, class_weight=None, coef0=0.0,\n",
       "    decision_function_shape='ovr', degree=3, gamma=0.3, kernel='rbf',\n",
       "    max_iter=-1, probability=False, random_state=None, shrinking=True,\n",
       "    tol=0.001, verbose=False)"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = svm.SVC(kernel='rbf', gamma=0.30, C=2000)\n",
    "clf.fit(pca_X_train, y_train_svm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [],
   "source": [
    "predictions = clf.predict(pca_X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Confustion matrix and results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.807\n",
      "F1 Score: 0.791\n",
      "Recall score: 0.733\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "confusion_matrix = metrics.confusion_matrix(y_test_svm, predictions)\n",
    "\n",
    "accuracy = metrics.accuracy_score(y_test_svm, predictions)\n",
    "\n",
    "print('Accuracy: {0:.3f}'.format(accuracy))\n",
    "print('F1 Score: {0:.3f}'.format(metrics.f1_score(y_test_svm, predictions)))\n",
    "print('Recall score: {0:.3f}'.format(metrics.recall_score(y_test_svm, predictions)))\n",
    "\n",
    "plt.figure(figsize=(8, 8))\n",
    "sns.heatmap(confusion_matrix, annot=True, fmt=\"d\", square=True, cmap=\"Blues\");\n",
    "plt.ylabel(\"Actual label\");\n",
    "plt.xlabel(\"Predicted label\");\n",
    "plt.title(f\"Accuracy Score: {accuracy:.3f}\", size = 15);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Conclusion\n",
    "CNNs perfrom better than classical SVN method in this dataset, but teh differenece is not that enormous. CNNs were strictly created for analyzing visual imagery, so I am positively surprised that SVM could result in over 80% accuracy. I think with proper tunning we could see "
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}