cnn网络实现

# -*- coding: utf-8 -*-

"""

Created on Sun Nov 12 14:10:16 2017

@author: jssyhhghf

"""

# CNN

import tensorflow as tf

from tensorflow.examples.tutorials.mnist import input_data

#载入数据集

mnist = input_data.read_data_sets(r"E:\anaconda\tensorflow\tensor_mnist-master\MNIST_data",one_hot=True)

#每个批次的大小

batch_size = 100

#计算一共有多少个批次

n_batch=mnist.train.num_examples//batch_size

#参数概要

def variable_summaries(var):

with tf.name_scope('summaries'):

mean = tf.reduce_mean(var)

tf.summary.scalar('mean',mean)#平均值

with tf.name_scope('stdder'):

stdder = tf.sqrt(tf.reduce_mean(tf.square(var-mean)))

tf.summary.scalar('stdder',stdder)        #标准差

tf.summary.scalar('max',tf.reduce_max(var))#最大值

tf.summary.scalar('min',tf.reduce_min(var))#最小值

tf.summary.histogram('histogram',var)      #直方图

#初始化权值

def weight_variable(shape,name):

initial = tf.truncated_normal(shape,stddev=0.1) #生成一个截断的正态分布

return tf.Variable(initial,name=name)

#初始化偏置

def bias_variable(shape,name):

initial = tf.constant(0.1,shape=shape)

return tf.Variable(initial,name=name)

#卷积层

def conv2d(x,W):

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

#x:tensor[batch,height,width,channels]

#W:卷积核[height,width,inchannels.outchannels]

#strides步长，第0和第3个都是1,1代表x方向的步长，2代表y方向的步长

#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')

with tf.name_scope('input'):

x=tf.placeholder(tf.float32,[None,784])

y=tf.placeholder(tf.float32,[None,10])

with tf.name_scope('x_image'):

#改变x的格式为4D向量

x_image = tf.reshape(x,[-1,28,28,1])

with tf.name_scope('Conv1'):

#初始化第一个卷基层的权值和偏置

with tf.name_scope('W_conv1'):

W_conv1=weight_variable([5,5,1,32],name='W_conv1')#5*5的采样窗口,32个卷积核从一个平面抽取特征

with tf.name_scope('b_conv1'):

b_conv1=bias_variable([32],name='b_conv1')#每一个卷积核一个偏置值

with tf.name_scope('conv2d_1'):

#把x_image和权值向量进行卷积，再加上偏置值，然后应用于relu激活函数

conv2d_1=conv2d(x_image,W_conv1)+b_conv1

with tf.name_scope('relu'):

h_conv1 = tf.nn.relu(conv2d_1)

with tf.name_scope('h_pool1'):

h_pool1 = max_pool_2x2(h_conv1)

with tf.name_scope('Conv2'):

with tf.name_scope('W_conv2'):

#初始化第二个卷基层的权值和偏置

W_conv2=weight_variable([5,5,32,64],name='W_conv2')#5*5的采样窗口,64个卷积核从32个平面抽取特征

with tf.name_scope('b_conv2'):

b_conv2=bias_variable([64],name='b_conv2')#每一个卷积核一个偏置值

with tf.name_scope('conv2d_2'):

conv2d_2=conv2d(h_pool1,W_conv2)+b_conv2

#把h_pool1和权值向量进行卷积，再加上偏置值，然后应用于relu激活函数

with tf.name_scope('relu'):

h_conv2 = tf.nn.relu(conv2d_2)

with tf.name_scope('h_pool2'):

h_pool2 = max_pool_2x2(h_conv2)

#28*28的图片第一次卷积后还是28*28,第一次池化为14*14

#第二次卷积后卫14*14，第二次池化为7*7

#经过上面的操作得到64张7*7的平面

with tf.name_scope('layer1'):

#初始化第一个全连接层的权值

with tf.name_scope('weight'):

W_fc1 = weight_variable([7*7*64,1024],name='W_fc1')

with tf.name_scope('bias'):

b_fc1 = bias_variable([1024],name='b_fc1')

with tf.name_scope('flat'):

#把池化层2的输出扁平化为1维

h_pool2_flat = tf.reshape(h_pool2,[-1,7*7*64],name='h_pool')

with tf.name_scope('relu'):

h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat,W_fc1)+b_fc1)

with tf.name_scope('keep_prob'):

#keep_prob用来表示神经元的输出概率

keep_prob = tf.placeholder(tf.float32,name='keep_prob')

with tf.name_scope('h_fc1_drop'):

h_fc1_drop=tf.nn.dropout(h_fc1,keep_prob,name='h_fc1_drop')

#初始化第二个全连接层

with tf.name_scope('layer2'):

with tf.name_scope('weight'):

W_fc2 = weight_variable([1024,10],name='W_fc2')

with tf.name_scope('bias'):

b_fc2 = bias_variable([10],name='b_fc2')

with tf.name_scope('softmax'):

prediction = tf.nn.softmax(tf.matmul(h_fc1_drop,W_fc2)+b_fc2)

with tf.name_scope('cross_entropy'):

#交叉熵代价函数

cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y,logits=prediction),name='cross_entropy')

tf.summary.scalar('cross_entropy',cross_entropy)

with tf.name_scope('train'):

#使用Adamoption进行优化

train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)

with tf.name_scope('accuracy'):

with tf.name_scope('correct_prediction'):

correct_prediction = tf.equal(tf.argmax(prediction,1),tf.argmax(y,1))

with tf.name_scope('accuracy'):

accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32))

tf.summary.scalar('accuracy',accuracy)

megred = tf.summary.merge_all()

with tf.Session() as sess:

sess.run(tf.global_variables_initializer())

train_writer = tf.summary.FileWriter(r'E:\anaconda\tensorflow\logs\train',sess.graph)

test_writer = tf.summary.FileWriter(r'E:\anaconda\tensorflow\logs\test',sess.graph)

for i in range(1001):

batch_xs,batch_ys = mnist.train.next_batch(batch_size)

sess.run(train_step, feed_dict={x:batch_xs,y:batch_ys,keep_prob:0.5})

summary=sess.run(megred,feed_dict={x:batch_xs,y:batch_ys,keep_prob:1.0})

train_writer.add_summary(summary,i)

batch_xs,batch_ys = mnist.test.next_batch(batch_size)

summary=sess.run(megred,feed_dict={x:batch_xs,y:batch_ys,keep_prob:1.0})

test_writer.add_summary(summary,i)

if (i%100==0):

test_acc = sess.run(accuracy,feed_dict={x:mnist.test.images,y:mnist.test.labels,keep_prob:1.0})

train_acc = sess.run(accuracy,feed_dict={x:mnist.train.images,y:mnist.train.labels,keep_prob:1.0})

print ('Iter'+str(i)+', Testing Accuracy='+str(test_acc)+',Training Accuracy='+str(train_acc))