卷积神经网络：代码实现

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt
from tensorflow.examples.tutorials.mnist import input_data

mnist = input_data.read_data_sets('data/', one_hot=True)
trainimg   = mnist.train.images
trainlabel = mnist.train.labels
testimg    = mnist.test.images
testlabel  = mnist.test.labels

# 输入和输出 
n_input  = 784 
n_output = 10

#卷积神经网络的参数初始化（w，b）
weights  = {
        'wc1': tf.Variable(tf.random_normal([3, 3, 1, 64], stddev=0.1)), #第一层卷积层权重参数[3, 3, 1, 64]卷积核的大小（3*3*1）；卷积核的个数64(特征图的个数)
        'wc2': tf.Variable(tf.random_normal([3, 3, 64, 128], stddev=0.1)), #第二层卷积层权重参数[3, 3, 64, 128]卷积核的大小（3*3*64（与输入图像深度对应））；卷积核的个数128(特征图的个数)
        'wd1': tf.Variable(tf.random_normal([7*7*128, 1024], stddev=0.1)),#第一层全连接层权重参数（由于该模型中卷积并未改变输入图像的大小，经过两次池化原始图像大小（28*28）变为（7*7））
        'wd2': tf.Variable(tf.random_normal([1024, n_output], stddev=0.1))#第二层全连接层权重参数（10分类）
    }
biases   = {
        'bc1': tf.Variable(tf.random_normal([64], stddev=0.1)),
        'bc2': tf.Variable(tf.random_normal([128], stddev=0.1)),
        'bd1': tf.Variable(tf.random_normal([1024], stddev=0.1)),
        'bd2': tf.Variable(tf.random_normal([n_output], stddev=0.1))
    }
#卷积层定义
def conv_basic(_input, _w, _b, _keepratio):
        # 输入预处理（转换为TensorFlow支持的格式）的
        _input_r = tf.reshape(_input, shape=[-1, 28, 28, 1])#第一维：batchsize的大小（-1让TensorFlow根据其余值推断该值的大小）；第二维：图像的高度；第三维：图像的宽度；第四维：图像的深度
        
        # 第一层卷积
        _conv1 = tf.nn.conv2d(_input_r, _w['wc1'], strides=[1, 1, 1, 1], padding='SAME')
        #print（help（tf.nn.conv2d））查看函数的帮助文档
        #strides=[batchsize的stride大小, h的stride大小, w的stride大小, c的stride大小]
        #padding='SAME'/'VALID':自动填充0（推荐）/不进行填充
        #_mean, _var = tf.nn.moments(_conv1, [0, 1, 2])
        #_conv1 = tf.nn.batch_normalization(_conv1, _mean, _var, 0, 1, 0.0001)
        
        _conv1 = tf.nn.relu(tf.nn.bias_add(_conv1, _b['bc1']))#卷积之后进行激活
        
        _pool1 = tf.nn.max_pool(_conv1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
        #池化操作，ksize窗口大小（batchsize的大小；图像的高度；图像的宽度；图像的深度），strides=[1, 2, 2, 1]：h和w方向步长均为2
       
        _pool_dr1 = tf.nn.dropout(_pool1, _keepratio)#dropout（随机地减少部分节点）
        
        # 第二层卷积
        _conv2 = tf.nn.conv2d(_pool_dr1, _w['wc2'], strides=[1, 1, 1, 1], padding='SAME')
        #_mean, _var = tf.nn.moments(_conv2, [0, 1, 2])
        #_conv2 = tf.nn.batch_normalization(_conv2, _mean, _var, 0, 1, 0.0001)
        _conv2 = tf.nn.relu(tf.nn.bias_add(_conv2, _b['bc2']))
        _pool2 = tf.nn.max_pool(_conv2, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')
        _pool_dr2 = tf.nn.dropout(_pool2, _keepratio)
       
       # 全连接层
        _dense1 = tf.reshape(_pool_dr2, [-1, _w['wd1'].get_shape().as_list()[0]])#定义全连接的输入
        # 第一层全连接层（神经网络）
        _fc1 = tf.nn.relu(tf.add(tf.matmul(_dense1, _w['wd1']), _b['bd1']))
        _fc_dr1 = tf.nn.dropout(_fc1, _keepratio)
        # 第一、二层全连接层
        _out = tf.add(tf.matmul(_fc_dr1, _w['wd2']), _b['bd2'])
        # 定义返回值
        out = { 'input_r': _input_r, 'conv1': _conv1, 'pool1': _pool1, 'pool1_dr1': _pool_dr1,
            'conv2': _conv2, 'pool2': _pool2, 'pool_dr2': _pool_dr2, 'dense1': _dense1,
            'fc1': _fc1, 'fc_dr1': _fc_dr1, 'out': _out
        }
        return out

x = tf.placeholder(tf.float32, [None, n_input])
y = tf.placeholder(tf.float32, [None, n_output])
keepratio = tf.placeholder(tf.float32)



_pred = conv_basic(x, weights, biases, keepratio)['out']
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits_v2(y, _pred))
optm = tf.train.AdamOptimizer(learning_rate=0.001).minimize(cost)
_corr = tf.equal(tf.argmax(_pred,1), tf.argmax(y,1)) 
accr = tf.reduce_mean(tf.cast(_corr, tf.float32)) 
init = tf.global_variables_initializer()
    
sess = tf.Session()
sess.run(init)

training_epochs = 10
batch_size      = 16  #网络结果比较复杂，这里取小一些，方便演示，正常情况下要稍大一些
display_step    = 1
for epoch in range(training_epochs):
    avg_cost = 0.
    #total_batch = int(mnist.train.num_examples/batch_size)
    total_batch = 10 #简单示例，正常情况如上
    # Loop over all batches
    for i in range(total_batch):
        batch_xs, batch_ys = mnist.train.next_batch(batch_size)
        # Fit training using batch data
        sess.run(optm, feed_dict={x: batch_xs, y: batch_ys, keepratio:0.7})
        # Compute average loss
        avg_cost += sess.run(cost, feed_dict={x: batch_xs, y: batch_ys, keepratio:1.})/total_batch

    # Display logs per epoch step
    if epoch % display_step == 0: 
        print ("Epoch: %03d/%03d cost: %.9f" % (epoch, training_epochs, avg_cost))
        train_acc = sess.run(accr, feed_dict={x: batch_xs, y: batch_ys, keepratio:1.})
        print (" Training accuracy: %.3f" % (train_acc))
        #test_acc = sess.run(accr, feed_dict={x: testimg, y: testlabel, keepratio:1.})
        #print (" Test accuracy: %.3f" % (test_acc))

运行结果：

可以看出，损失值在不断减少，训练集的精度也有改善