Tensorflow卷积神经网络入门练习

注以下代码在CPU版本的Tensorflow并不容易跑起来

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

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

# 每个批次的大小
batch_size = 100
# 一共有多少批次
n_batch = mnist.train.num_examples // batch_size


# 初始化权值
def weight_variable(shape):
    initial = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(initial)


# 初始化偏置
def bias_variable(shape):
    initial = tf.constant(0.1, shape=shape)
    return tf.Variable(initial)


# 卷积层
def conv2d(x, W):
    # W[filter_height, filter_width, in_channels, out_channels]
    # filter_height 卷积核高度 filter_width 卷积核宽度 in_channels 输入通道数 out_channels 输出通道数
    # strider[0] = strider[3] = 1
    # strider[1] 代表x方向步长
    # strider[2] 代表y方向步长
    # padding = "SAME" "VALID"

    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding="SAME")


# 池化层
def max_pool_2x2(x):
    # ksize (1,x,y,1) 代表池化层的大小
    # strides代表步长，参数设置与conv2d相同
    return tf.nn.max_pool(x, ksize=[1, 2, 2,1],strides=[1, 2, 2, 1],padding="SAME")

# 定义两个placeholder
x = tf.placeholder(tf.float32, [None, 28*28])
y = tf.placeholder(tf.float32, [None, 10])

# 改变图片的输入格式
# [batch, in_height, in_weight, in_channels]
# -1代表批次任意
x_image = tf.reshape(x,[-1, 28, 28, 1])

# 初始化第一个卷积核的权值和偏置
W_conv1 = weight_variable([5,5,1,32])   # 5*5的卷积窗口,32个卷积核从1个平面提取特征
b_conv1 = bias_variable([32]) # 每个卷积核一个偏置

# 进行第一层卷积操作,并进行relu激活函数，在进行2*2最大池化
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1) # 每个卷积核一个偏置

# 初始化第二个卷积核的权值和偏置
W_conv2 = weight_variable([5, 5, 32, 64])   # 5*5的卷积窗口，64个卷积核从32个平面提取特征
b_conv2 = bias_variable([64])

# 进行第二层卷积操作并加偏置和激活函数
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2)

# 28*28的图片第一次卷积后还是28*28，第一次池化后变为14*14
# 第二次卷积之后还是14*14， 第二次池化后变为7*7
# 经上述操作后变为64张7*7的图片

# 初始化第一个全连接层
W_fc1 = weight_variable([7*7*64, 1024])  # 全连接层的第一层有1024个
b_fc1 = bias_variable([1024])

# 把池化层的输出一维化
h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
# 第一个全连接层输入并过relu激活函数
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

# keep_prob表示神经元输出的概率
keep_prob = tf.placeholder(tf.float32)
# 对第一个神经元进行dropout操作
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])

# 计算输出,使用softmax
prediction = tf.nn.softmax(tf.matmul(h_fc1_drop, W_fc2)+b_fc2)

# 交叉熵代价函数
cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y, logits=prediction))
# 使用AdamOptimizer优化器训练
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
# 结果存放在一个布尔列表中
correct_prediction = tf.equal(tf.argmax(prediction,1),tf.argmax(y,1))
# 求准确率
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

with tf.Session() as sess:
    # 初始化
    sess.run(tf.global_variables_initializer())
    for epoch in range(21):
        for batch in range(n_batch):
            batch_xs, batch_ys = mnist.train.next_batch(batch_size)
            sess.run(train_step,feed_dict={x:batch_xs,y:batch_ys,keep_prob:0.7})

        acc = sess.run(accuracy, feed_dict={x:mnist.test.images, y:mnist.test.labels,keep_prob:1.0})
        print("Iter " + str(epoch) + ", Testing Accuracy= " + str(acc))