Day93-CNN_Brief_HW.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 作業目標:\n",
    "    運用 Keras 模組建構CNN, 了解 CNN 的架構\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 作業重點\n",
    "(1)嘗試比對 Dense 與 layers.Conv2D 架構NN 的差異\n",
    "\n",
    "(2) 有沒有Pooling layer, 對於參數量的差異\n",
    "\n",
    "注意: input_shape 請勿修改"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "2.4.3\n"
    }
   ],
   "source": [
    "#導入相關模組\n",
    "import keras\n",
    "from keras import layers\n",
    "from keras import models\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D, Activation, MaxPooling2D, Flatten, Dense\n",
    "\n",
    "#確認keras 版本\n",
    "print(keras.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "Model: \"sequential_2\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\ndense_2 (Dense)              (None, 28, 28, 100)       200       \n_________________________________________________________________\nactivation_2 (Activation)    (None, 28, 28, 100)       0         \n_________________________________________________________________\ndense_3 (Dense)              (None, 28, 28, 10)        1010      \n_________________________________________________________________\nactivation_3 (Activation)    (None, 28, 28, 10)        0         \n=================================================================\nTotal params: 1,210\nTrainable params: 1,210\nNon-trainable params: 0\n_________________________________________________________________\n"
    }
   ],
   "source": [
    "'''\n",
    "model = Sequential([\n",
    "    Dense(xxx),\n",
    "    Activation(xxx),\n",
    "    Dense(xxx),\n",
    "    Activation(xxx),\n",
    "])\n",
    "'''\n",
    "model=Sequential([\n",
    "    Dense(units=100,input_shape=(28,28,1)),\n",
    "    Activation('relu'),\n",
    "    Dense(units=10),\n",
    "    Activation('softmax'),\n",
    "])\n",
    "\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "attachments": {
    "CNN_Model.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# layers.Conv2D 模型, 用作比對\n",
    "\n",
    "![CNN_Model.png](attachment:CNN_Model.png)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": "Model: \"sequential_1\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nconv2d (Conv2D)              (None, 26, 26, 32)        320       \n_________________________________________________________________\ndense (Dense)                (None, 26, 26, 100)       3300      \n_________________________________________________________________\nactivation (Activation)      (None, 26, 26, 100)       0         \n_________________________________________________________________\ndense_1 (Dense)              (None, 26, 26, 10)        1010      \n_________________________________________________________________\nactivation_1 (Activation)    (None, 26, 26, 10)        0         \n=================================================================\nTotal params: 4,630\nTrainable params: 4,630\nNon-trainable params: 0\n_________________________________________________________________\n"
    }
   ],
   "source": [
    "'''\n",
    "#建立一個序列模型\n",
    "model = models.Sequential()\n",
    "#建立兩個卷績層, 32 個內核, 內核大小 3x3, \n",
    "#輸入影像大小 28x28x1\n",
    "\n",
    "#建立一個全連接層\n",
    "model.add(Dense(units=100))\n",
    "model.add(Activation('relu'))\n",
    "\n",
    "#建立一個輸出層, 並採用softmax\n",
    "model.add(Dense(units=10))\n",
    "model.add(Activation('softmax'))\n",
    "'''\n",
    "model=models.Sequential()\n",
    "\n",
    "model.add(layers.Conv2D(32, (3,3), input_shape =(28,28,1)))\n",
    "model.add(Dense(units=100))\n",
    "model.add(Activation('relu'))\n",
    "model.add(Dense(units=10))\n",
    "model.add(Activation('softmax'))\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}