tensorflow的hellow world：mnist手写识别

#加载数据
from tensorflow.examples.tutorials.mnist import input_data
mnist=input_data.read_data_sets("MNIST_data/",one_hot=True)


import tensorflow as tf


#变量定义
x=tf.placeholder("float",[None,784])
W=tf.Variable(tf.zeros([784,10]))
b=tf.Variable(tf.zeros([10]))


#设置模型
y=tf.nn.softmax(tf.matmul(x,W)+b)
y_=tf.placeholder("float",[None,10])
cross_entropy=-tf.reduce_sum(y_*tf.log(y))
train_step=tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)


#开始训练
init=tf.global_variables_initializer()
sess=tf.Session()
sess.run(init)
for i in range(1000):
    batch_xs,batch_ys=mnist.train.next_batch(100)
    sess.run(train_step,feed_dict={x:batch_xs,y_:batch_ys})


#针对测试集进行模型评估
correct_prediction=tf.equal(tf.argmax(y,1),tf.argmax(y_,1))
accuracy=tf.reduce_mean(tf.cast(correct_prediction,"float"))
print(sess.run(accuracy,feed_dict={x:mnist.test.images,y_:mnist.test.labels}))