HW2_YangHaoqing

.ipynb_checkpoints/keras-checkpoint.ipynb

@@ -0,0 +1,199 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Keras Tutorial"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Keras is a popular deep learning framework and it's easy to get started."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import keras\n",
+    "from keras.datasets import mnist\n",
+    "from keras.models import Sequential\n",
+    "from keras.layers import Dense,Activation, Dropout, Flatten, Conv2D, MaxPooling2D\n",
+    "\n",
+    "BATCH_SIZE = 128\n",
+    "NUM_CLASSES = 10\n",
+    "NUM_EPOCHS = 10\n",
+    "nb_filters = 32\n",
+    "pool_size = (2,2)\n",
+    "kernel_size = (3,3)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "First, we read the mnist data and preprocess them."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# download and load the data (split them between train and test sets)\n",
+    "(x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
+    "\n",
+    "# expand the channel dimension\n",
+    "x_train = x_train.reshape(x_train.shape[0], 28, 28, 1)\n",
+    "x_test = x_test.reshape(x_test.shape[0], 28, 28, 1)\n",
+    "input_shape = (28, 28, 1)\n",
+    "\n",
+    "# make the value of pixels from [0, 255] to [0, 1] for further process\n",
+    "x_train = x_train.astype('float32') / 255.\n",
+    "x_test = x_test.astype('float32') / 255.\n",
+    "\n",
+    "# convert class vectors to binary class matrics\n",
+    "y_train = keras.utils.to_categorical(y_train, NUM_CLASSES)\n",
+    "y_test = keras.utils.to_categorical(y_test, NUM_CLASSES)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Then, we define the model, object function and optimizer that we use to classify."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(60000, 28, 28, 1) (60000, 10) (10000, 28, 28, 1) (10000, 10)\n",
+      "WARNING:tensorflow:From /home/luna/.local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
+      "\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(x_train.shape,y_train.shape,x_test.shape,y_test.shape)\n",
+    "\n",
+    "# define the model\n",
+    "model = Sequential()\n",
+    "model.add(Conv2D(nb_filters,kernel_size,input_shape=input_shape,padding='valid'))  \n",
+    "model.add(Activation('relu'))\n",
+    "model.add(Conv2D(nb_filters,kernel_size))\n",
+    "model.add(Activation('relu'))\n",
+    "model.add(MaxPooling2D(pool_size=pool_size))\n",
+    "model.add(Dropout(0.25))\n",
+    "model.add(Flatten())\n",
+    "model.add(Dense(128))\n",
+    "model.add(Activation('relu'))\n",
+    "model.add(Dropout(0.5))\n",
+    "model.add(Dense(NUM_CLASSES))\n",
+    "model.add(Activation('softmax'))\n",
+    "\n",
+    "# define the object function, optimizer and metrics\n",
+    "model.compile(loss = 'categorical_crossentropy',optimizer='adadelta',metrics=['accuracy'])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Next, we can start to train and evaluate!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "WARNING:tensorflow:From /home/luna/.local/lib/python3.6/site-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
+      "\n",
+      "Train on 60000 samples, validate on 10000 samples\n",
+      "Epoch 1/10\n",
+      "60000/60000 [==============================] - 87s 1ms/step - loss: 0.2938 - accuracy: 0.9104 - val_loss: 0.0648 - val_accuracy: 0.9800\n",
+      "Epoch 2/10\n",
+      "60000/60000 [==============================] - 82s 1ms/step - loss: 0.1007 - accuracy: 0.9706 - val_loss: 0.0467 - val_accuracy: 0.9842\n",
+      "Epoch 3/10\n",
+      "60000/60000 [==============================] - 80s 1ms/step - loss: 0.0759 - accuracy: 0.9767 - val_loss: 0.0386 - val_accuracy: 0.9872\n",
+      "Epoch 4/10\n",
+      "60000/60000 [==============================] - 88s 1ms/step - loss: 0.0639 - accuracy: 0.9814 - val_loss: 0.0373 - val_accuracy: 0.9880\n",
+      "Epoch 5/10\n",
+      "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0557 - accuracy: 0.9837 - val_loss: 0.0313 - val_accuracy: 0.9891\n",
+      "Epoch 6/10\n",
+      "60000/60000 [==============================] - 85s 1ms/step - loss: 0.0502 - accuracy: 0.9847 - val_loss: 0.0330 - val_accuracy: 0.9896\n",
+      "Epoch 7/10\n",
+      "60000/60000 [==============================] - 84s 1ms/step - loss: 0.0481 - accuracy: 0.9857 - val_loss: 0.0303 - val_accuracy: 0.9900\n",
+      "Epoch 8/10\n",
+      "60000/60000 [==============================] - 82s 1ms/step - loss: 0.0434 - accuracy: 0.9870 - val_loss: 0.0304 - val_accuracy: 0.9906\n",
+      "Epoch 9/10\n",
+      "60000/60000 [==============================] - 82s 1ms/step - loss: 0.0391 - accuracy: 0.9881 - val_loss: 0.0271 - val_accuracy: 0.9909\n",
+      "Epoch 10/10\n",
+      "60000/60000 [==============================] - 82s 1ms/step - loss: 0.0366 - accuracy: 0.9888 - val_loss: 0.0311 - val_accuracy: 0.9901\n",
+      "60000/60000 [==============================] - 17s 288us/step\n",
+      "Training loss: 0.0156, Training accuracy: 99.50%\n",
+      "10000/10000 [==============================] - 3s 285us/step\n",
+      "Testing loss: 0.0311, Testing accuracy: 99.01%\n"
+     ]
+    }
+   ],
+   "source": [
+    "# train\n",
+    "model.fit(x_train,y_train,batch_size=BATCH_SIZE,epochs=NUM_EPOCHS,validation_data=(x_test,y_test))\n",
+    "\n",
+    "# evaluate\n",
+    "score_train,acc_train = model.evaluate(x_train,y_train,batch_size=BATCH_SIZE)\n",
+    "print('Training loss: %.4f, Training accuracy: %.2f%%' % (score_train,acc_train*100))\n",
+    "score_test,acc_test = model.evaluate(x_test,y_test,batch_size=BATCH_SIZE)\n",
+    "print('Testing loss: %.4f, Testing accuracy: %.2f%%' % (score_test,acc_test*100))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Q6:\n",
+    "Please print the training and testing accuracy."
+   ]
+  }
+ ],
+ "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": "3.6.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}