Tensorflow CNN 的简单实现

教学视频：莫烦python

有空自己整理复写一遍~mark

# encoding:utf8

import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

mnist = input_data.read_data_sets('MNIST_data', one_hot=True)


def compute_accuracy(v_xs, v_ys):
    global prediction
    y_pre = sess.run(prediction, feed_dict={xs: v_xs, keep_prob: 1})
    correct_predition = tf.equal(tf.argmax(y_pre, 1), tf.argmax(v_ys, 1))
    accuracy = tf.reduce_mean(tf.cast(correct_predition, tf.float32))  # tf.cast 为数据和转换函数
    result = sess.run(accuracy, feed_dict={xs: v_xs, ys: v_ys, keep_prob: 1})
    return result


def weight_variable(shape):
    initial = tf.truncated_normal(shape, stddev=0.1)#tf.truncated_normal生成截断正太分布
    return tf.Variable(initial)  #定义变量


def bias_variable(shape):
    initial = tf.constant(0.1, shape=shape)
    return tf.Variable(initial)


## 定义一个卷积层 输入为x,w为权重
def conv2d(x, W):
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
## padding has another choice "VALID"

## 定义池化层
def max_pool_2x2(x):
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
# 池化x,ksize[batch,height,width,channels]为池化窗口的大小,不在batch和channels上池化,取值为1
# strides为窗口在每个维度滑动的步长

# define placeholder for inputs to network
xs = tf.placeholder(tf.float32, [None, 784])/255
ys = tf.placeholder(tf.float32, [None, 10])
keep_prob = tf.placeholder(tf.float32)
x_image = tf.reshape(xs, [-1, 28, 28, 1])
# 将输入的变量变为一个4维向量
# print x_image.shape


## conv1 layer ##
W_conv1 = weight_variable([5, 5, 1, 32])  # 卷积核的大小为5x5,1为通道数,32为传入下一个神经元的节点数
b_conv1 = bias_variable([32])
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)

## conv2 layer ##
W_conv2 = W_conv1 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)

## fc1 layer ##
W_fc1 = weight_variable([7 * 7 * 64, 1024])
b_fc1 = bias_variable([1024])
h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
h_fc1_dropout = tf.nn.dropout(h_fc1, keep_prob)

## fc2 layer ##
W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])
prediction = tf.nn.softmax(tf.matmul(h_fc1_dropout, W_fc2) + b_fc2)

## op ##

cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys * tf.log(prediction), reduction_indices=[1]))

#reduction_indices 在哪一维上求解
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
# 选择Adam ,learning rate:0.0001

sess = tf.Session()
init = tf.global_variables_initializer()
sess.run(init)

for i in range(1000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={xs: batch_xs, ys: batch_ys, keep_prob: 0.5})
    if i % 50 == 0:
        print compute_accuracy(mnist.test.images[:1000], mnist.test.labels[:1000])