MNIST数据集分类简单版本

 

import tensorflow as tf

from tensorflow.examples.tutorials.mnist import input_data

#载入数据集

mnist = input_data.read_data_sets("/data/stu05/mnist_data",one_hot=True)

 

 

Extracting /data/stu05/mnist_data/train-images-idx3-ubyte.gz
Extracting /data/stu05/mnist_data/train-labels-idx1-ubyte.gz
Extracting /data/stu05/mnist_data/t10k-images-idx3-ubyte.gz
Extracting /data/stu05/mnist_data/t10k-labels-idx1-ubyte.gz

 

#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

​

#定义两个placeholder,None=100,28*28=784,即100行，784列

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

#0-9个输出标签

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

​

#创建一个简单的神经网络，只有输入层和输出层

W = tf.Variable(tf.zeros([784,10]))

b = tf.Variable(tf.zeros([1,10]))

#softmax函数转化为概率值

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

​

#二次代价函数

loss = tf.reduce_mean(tf.square(y-prediction))

#使用梯度下降法

train_step = tf.train.GradientDescentOptimizer(0.2).minimize(loss)

​

#初始化变量

init = tf.global_variables_initializer()

​

#tf.equal()比较函数大小是否相同，相同为True,不同为false;tf.argmax()：求y=1在哪个位置，求概率最大在哪个位置

#argmax返回一维张量中最大的值所在的位置，结果存放在一个布尔型列表中

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

#求准确率

#cast转化类型，将布尔型转化为32位浮点型,True=1.0,False=0.0;再求平均值

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

​

with tf.Session() as sess:

    sess.run(init)

    #将所有图片训练21次

    for epoch in range(21):

        #训练一次所有的图片

        for batch in range(n_batch):

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

            #feed_dict传入训练集的图片和标签

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

        #传入测试集的图片和标签

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

        print("Iter"+str(epoch)+",Testing Accuracy："+str(acc))

 

 

 

Iter0,Testing Accuracy：0.8303
Iter1,Testing Accuracy：0.8708
Iter2,Testing Accuracy：0.8821
Iter3,Testing Accuracy：0.8885
Iter4,Testing Accuracy：0.8941
Iter5,Testing Accuracy：0.8973
Iter6,Testing Accuracy：0.9001
Iter7,Testing Accuracy：0.9013
Iter8,Testing Accuracy：0.9038
Iter9,Testing Accuracy：0.9048
Iter10,Testing Accuracy：0.9068
Iter11,Testing Accuracy：0.9068
Iter12,Testing Accuracy：0.9084
Iter13,Testing Accuracy：0.9094
Iter14,Testing Accuracy：0.9097
Iter15,Testing Accuracy：0.9107
Iter16,Testing Accuracy：0.9118
Iter17,Testing Accuracy：0.9116
Iter18,Testing Accuracy：0.9127
Iter19,Testing Accuracy：0.9136
Iter20,Testing Accuracy：0.9146

 

 

 

 

 

​