1、RNN（Recurrent Neural Network）循环神经网络模型
详见RNN循环神经网络：https://www.cnblogs.com/pinard/p/6509630.html
2、LSTM（Long Short Term Memory）长短期记忆神经网络模型
详见LSTM长短期记忆神经网络：http://www.cnblogs.com/pinard/p/6519110.html
3、LSTM长短期记忆神经网络处理Mnist数据集
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 | import tensorflow as tffrom tensorflow.examples.tutorials.mnist import input_datafrom tensorflow.contrib import rnn# 载入数据集mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)# 输入图片是28*28n_inputs = 28 # 输入一行，一行有28个数据（28个像素点），即输入序列长度为28max_time = 28 # 一共28行lstm_size = 100 # 隐层单元n_classes = 10 # 10个分类batch_size = 50 # 每批次50个样本n_batch = mnist.train.num_examples // batch_size # 计算一共有多少个批次# 这里的none表示第一个维度可以是任意的长度x = tf.placeholder(tf.float32, [None, 784])# 正确的标签y = tf.placeholder(tf.float32, [None, 10])# 初始化权值weights = tf.Variable(tf.truncated_normal([lstm_size, n_classes], stddev=0.1))# 初始化偏置值biases = tf.Variable(tf.constant(0.1, shape=[n_classes]))# 定义RNN网络def RNN(X, weights, biases): inputs = tf.reshape(X, [-1, max_time, n_inputs]) # 定义LSTM基本CELL lstm_cell = rnn.BasicLSTMCell(lstm_size) # final_state[0]是cell state # final_state[1]是hidden_state outputs, final_state = tf.nn.dynamic_rnn(lstm_cell, inputs, dtype=tf.float32) results = tf.nn.softmax(tf.matmul(final_state[1], weights) + biases) return results# 计算RNN的返回结果prediction = RNN(x, weights, biases)# 损失函数cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=prediction, labels=y))# 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型列表中correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(prediction, 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)) # 把correct_prediction变为float32类型# 初始化init = tf.global_variables_initializer()with tf.Session() as sess: sess.run(init) for epoch in range(21): 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}) acc = sess.run(accuracy, feed_dict={x: mnist.test.images, y: mnist.test.labels}) print("Iter " + str(epoch) + ", Testing Accuracy= " + str(acc)) |
结果为：