1.MNIST数据库下载好后，在tensorflow/examples/tutorials/mnist/下建立文件夹MNIST_data即可运行本程序
2.关键在与理解Operation，Tensor，Graph，只有执行session.run()时操作才真正执行
import tensorflow.examples.tutorials.mnist.input_data as input_data mnist = input_data.read_data_sets(‘MNIST_data/‘,one_hot = True) import tensorflow as tf # 定义计算图 # 操作Operation为图的节点(如下面的tf.placeholder(tf.float32,[None,784])等) # 数据Tensor为图的边(如下面的x,y等) # 添加Operation时不会立即执行,只有执行session.run(Operation或Tensor)时才会真正执行 x = tf.placeholder(tf.float32,[None,784]) y = tf.placeholder(tf.float32,[None,10]) W = tf.Variable(tf.zeros([784,10]),tf.float32) b = tf.Variable(tf.zeros([10]),tf.float32) py = tf.nn.softmax(tf.matmul(x,W) + b) loss = -tf.reduce_mean(y*tf.log(py)) # 添加计算图节点global_variables_initializer(),返回初始化变量的Operation # 官方文档解释: Returns an Op that initializes global variables. init = tf.global_variables_initializer(); # 获得Session对象 sess = tf.Session() # 真正执行初始化节点init sess.run(init) # 训练MNIST数据库 # 添加train_step计算节点，这个计算节点完成梯度下降功能，train_step为一个Operation train_step = tf.train.GradientDescentOptimizer(0.01).minimize(loss) for i in range(10000): batch_xs,batch_ys = mnist.train.next_batch(1000) # 执行梯度下降节点，tensorflow会根据计算图反推依赖，提前计算依赖节点 # 由于x,y中含有None，需要feed_dict = {x:batch_xs,y:batch_ys}填充数据 sess.run(train_step,feed_dict = {x:batch_xs,y:batch_ys}) # observe gradient descent in training set if i%100 == 0: # 计算节点correct_prediction correct_prediction = tf.equal(tf.argmax(y,1),tf.argmax(py,1)) # 计算节点accuracy accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32)) # 反推图依赖，得到正确的accuracy print(‘training set accuracy: ‘,sess.run(accuracy,feed_dict={x:batch_xs,y:batch_ys})) # 观察测试集的performance correct_prediction = tf.equal(tf.argmax(y,1),tf.argmax(py,1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction,tf.float32)) print(‘test set accuracy: ‘,sess.run(accuracy,feed_dict={x:batch_xs,y:batch_ys})) 原文：https://www.cnblogs.com/liuzhan709/p/9346422.html