Tensorflow学习教程------利用卷积神经网络对mnist数据集进行分类_训练模型

Posted 2021-01-20 shuimuqingyang

原理就不多讲了，直接上代码，有详细注释。

#coding:utf-8

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

n_batch = mnist.train._num_examples // batch_size

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

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

#卷基层
def conv2d(x,W):
    return tf.nn.conv2d(x,W,strides=[1,1,1,1],padding=‘SAME‘)
#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‘)
#定义两个placeholder
x = tf.placeholder(tf.float32, [None,784])
y = tf.placeholder(tf.float32,[None,10])

#改变x的格式转为4D的向量[batch,in_height,in_width,in_channels]
x_image = tf.reshape(x, [-1,28,28,1])

#初始化第一个卷基层的权值和偏置
W_conv1 = weight_variable([5,5,1,32]) #5*5的采样窗口 32个卷积核从一个平面抽取特征 32个卷积核是自定义的
b_conv1 = bias_variable([32])  #每个卷积核一个偏置值

#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
h_conv1 = tf.nn.relu(conv2d(x_image,W_conv1)+b_conv1)
h_pool1 = max_pool_2x2(h_conv1) #进行max-pooling

#初始化第二个卷基层的权值和偏置
W_conv2 = weight_variable([5,5,32,64]) # 5*5的采样窗口 64个卷积核从32个平面抽取特征  由于前一层操作得到了32个特征图
b_conv2 = bias_variable([64]) #每一个卷积核一个偏置值

#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
h_conv2 = tf.nn.relu(conv2d(h_pool1,W_conv2) + b_conv2)
h_pool2 = max_pool_2x2(h_conv2) #进行max-pooling

#28x28的图片第一次卷积后还是28x28 第一次池化后变为14x14
#第二次卷积后 变为14x14 第二次池化后变为7x7
#通过上面操作后得到64张7x7的平面

#初始化第一个全连接层的权值
W_fc1 = weight_variable([7*7*64,1024])#上一层有7*7*64个神经元，全连接层有1024个神经元
b_fc1 = bias_variable([1024]) #1024个节点

#把第二个池化层的输出扁平化为一维
h_pool2_flat = tf.reshape(h_pool2,[-1,7*7*64])
#求第一个全连接层的输出
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat,W_fc1)+b_fc1)

#keep_prob用来表示神经元的输出概率
keep_prob  = tf.placeholder(tf.float32)
h_fc1_drop = tf.nn.dropout(h_fc1,keep_prob)

#初始化第二个全连接层
W_fc2 = weight_variable([1024,10])
b_fc2 = bias_variable([10])

#计算输出
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进行优化
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)) #argmax返回一维张量中最大的值所在的位置
#求准确率
accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32))
saver = tf.train.Saver()
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    for epoch in range(50):
        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))
　　saver.save(sess,save_path=‘/home/xxx/logs/mnistmodel‘,global_step=1)#将训练出来的权重参数保存

结果

Iter 0, Testing Accuracy= 0.8517
Iter 1, Testing Accuracy= 0.9612
Iter 2, Testing Accuracy= 0.9769
Iter 3, Testing Accuracy= 0.9804
Iter 4, Testing Accuracy= 0.9832
Iter 5, Testing Accuracy= 0.9844
Iter 6, Testing Accuracy= 0.988
Iter 7, Testing Accuracy= 0.9882
Iter 8, Testing Accuracy= 0.9875
Iter 9, Testing Accuracy= 0.9889
Iter 10, Testing Accuracy= 0.9891
Iter 11, Testing Accuracy= 0.9897
Iter 12, Testing Accuracy= 0.9891
Iter 13, Testing Accuracy= 0.9897
Iter 14, Testing Accuracy= 0.9905
Iter 15, Testing Accuracy= 0.9913
Iter 16, Testing Accuracy= 0.9908
Iter 17, Testing Accuracy= 0.9909
Iter 18, Testing Accuracy= 0.9913
Iter 19, Testing Accuracy= 0.9915
Iter 20, Testing Accuracy= 0.9902
Iter 21, Testing Accuracy= 0.9899
Iter 22, Testing Accuracy= 0.9912
Iter 23, Testing Accuracy= 0.9911
Iter 24, Testing Accuracy= 0.9907
Iter 25, Testing Accuracy= 0.9918
Iter 26, Testing Accuracy= 0.9919
Iter 27, Testing Accuracy= 0.9916
Iter 28, Testing Accuracy= 0.9899
Iter 29, Testing Accuracy= 0.9924
Iter 30, Testing Accuracy= 0.9913
Iter 31, Testing Accuracy= 0.992
Iter 32, Testing Accuracy= 0.9927
Iter 33, Testing Accuracy= 0.9919
Iter 34, Testing Accuracy= 0.9922
Iter 35, Testing Accuracy= 0.9918
Iter 36, Testing Accuracy= 0.9932
Iter 37, Testing Accuracy= 0.9924
Iter 38, Testing Accuracy= 0.9917
Iter 39, Testing Accuracy= 0.9919
Iter 40, Testing Accuracy= 0.9933
Iter 41, Testing Accuracy= 0.9924
Iter 42, Testing Accuracy= 0.9926
Iter 43, Testing Accuracy= 0.9932
Iter 44, Testing Accuracy= 0.9922
Iter 45, Testing Accuracy= 0.9925
Iter 46, Testing Accuracy= 0.9928
Iter 47, Testing Accuracy= 0.9935
Iter 48, Testing Accuracy= 0.9922
Iter 49, Testing Accuracy= 0.9926

下面来一个简洁版本的

# load MNIST data
from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("Mnist_data/", one_hot=True)

# start tensorflow interactiveSession
import tensorflow as tf
sess = tf.InteractiveSession()
batch_size = 50

n_batch = mnist.train._num_examples // batch_size

# weight initialization
def weight_variable(shape):
    initial = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(initial)

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

# convolution
def conv2d(x, W):
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding=‘SAME‘)
# pooling
def max_pool_2x2(x):
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding=‘SAME‘)

# Create the model
# placeholder
x = tf.placeholder("float", [None, 784])
y_ = tf.placeholder("float", [None, 10])
# variables
W = tf.Variable(tf.zeros([784,10]))
b = tf.Variable(tf.zeros([10]))

y = tf.nn.softmax(tf.matmul(x,W) + b)

# first convolutinal layer
w_conv1 = weight_variable([5, 5, 1, 32])
b_conv1 = bias_variable([32])

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

h_conv1 = tf.nn.relu(conv2d(x_image, w_conv1) + b_conv1)
h_pool1 = max_pool_2x2(h_conv1)

# second convolutional layer
w_conv2 = 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)

# densely connected 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)

# dropout
keep_prob = tf.placeholder("float")
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

# readout layer
w_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])

y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, w_fc2) + b_fc2)

# train and evaluate the model
cross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))
train_step = tf.train.AdagradOptimizer(1e-4).minimize(cross_entropy)
correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
sess.run(tf.initialize_all_variables())

for i in range(20000):
    batch = mnist.train.next_batch(50)
    if i%100 == 0:
        train_accuracy = accuracy.eval(feed_dict={x:batch[0], y_:batch[1], keep_prob:1.0})
        print ("step %d, train accuracy %g" %(i, train_accuracy))
    train_step.run(feed_dict={x:batch[0], y_:batch[1], keep_prob:0.5})

print ("test accuracy %g" % accuracy.eval(feed_dict={x:mnist.test.images, y_:mnist.test.labels, keep_prob:1.0}))