Tensorflow入门之Mnist数字识别(简单神经网络)

# coding=utf-8
import tensorflow as tf
#导入mnist数据集
from tensorflow.examples.tutorials.mnist import input_data
#若数据集不存在则下载更新
mnist = input_data.read_data_sets('MNIST_data',one_hot = True)

#定义层
def add_layer(inputs,in_size,out_size,activation_function=None):
    Weights = tf.Variable(tf.random_normal([in_size,out_size]))
    biases = tf.Variable(tf.zeros([1,out_size]))
    Wx_plus_b = tf.matmul(inputs,Weights)+biases
    if activation_function is None:
        outputs = Wx_plus_b
    else:
        outputs = activation_function(Wx_plus_b)
    return outputs

#定义准确度函数
def compute_accuracy(v_xs,v_ys):
    global  prediction
    y_pre = sess.run(prediction,feed_dict={xs:v_xs})
    """
    tf.argmax(A,axis)
    axis为0表示对行进行操作(对每列中的不同行进行比较)，返回对应最大值的索引
    axis为1表示对列进行操作(对每行中的不同列进行比较)，返回对应最大值的索引
    tf.equal()
    判断两矩阵或向量中的个元素是否相等
    """
    correct_prediction = tf.equal(tf.argmax(y_pre,1),tf.argmax(v_ys,1))
    """
    tf.cast()
    转换数据类型
    """
    accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32))
    result = sess.run(accuracy,feed_dict={xs:v_xs,ys:v_ys})
    return result

#定义输入层的placeholder
xs = tf.placeholder(tf.float32,[None,784])
ys = tf.placeholder(tf.float32,[None,10])

#添加输出层
prediction = add_layer(xs,784,10,activation_function=tf.nn.softmax)

#计算交叉熵,计算公式为H(p.q)=-sum(p(x)logq(x)),其中p为期望输出，q为实际输出
cross_entropy = tf.reduce_mean(-tf.reduce_sum(ys*tf.log(prediction),reduction_indices=[1]))
train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)

sess = tf.Session()
init = tf.initialize_all_variables()
sess.run(init)

#训练
for i in range(1000):
    batch_xs,batch_ys = mnist.train.next_batch(50)
    sess.run(train_step,feed_dict={xs:batch_xs,ys:batch_ys})
    if i%50 ==0:
        print(compute_accuracy(mnist.test.images,mnist.test.labels))