深度学习基础系列：RNN

深度学习基础系列：RNN
RNN，循环神经网络，Recurrent Neural Network。RNN对于处理有序的数据很有效，预测序列化的数据

在传统的神经网络模型中，从输入层到隐含层再到输出层，层与层之间是全连接的。
一个输入对应一个输出，多个输入对应多个输出，但是这些输入之间、输出之间相互没有关联，如下图所示：

但是这种普通的神经网络对于很多关于时间序列的问题却无能无力。例如，你要预测句子的下一个单词是什么，一般需要用到前面的单词，因为一个句子中前后单词并不是独立的。RNN之所以称为循环神经网路，即一个序列当前的输出与前面的输出也有关。具体的表现形式为网络会对前面时刻的信息进行记忆并应用于当前输出的计算中，即隐藏层之间的节点不再无连接而是有连接的，并且隐藏层的输入不仅包括输入层的输出还包括上一时刻隐藏层的输出。RNN的基本机构如下图所示：

普通的RNN，如果要预测的序列是一个很长的序列，则反向传播过程中存在梯度消失和梯度爆炸现象。
为了解决上述问题，提出了 LSTM RNN，Long Short-Term Memory，长短期记忆网络

LSTM的改进：增加了三个控制器——输入控制、输出控制、忘记控制


输入：考虑要不要将分线剧情加入到主线剧情，如果某些分线剧情比较重要，那么就会按重要程度，将其写入总线剧情，再进行分析。
忘记：如果分线剧情改变了我们对主线剧情的认知，那么忘记剧情就会对之前的剧情进行忘记，按比例替换为现在的新剧情。
所以主线剧情的更新就取决于输入控制和忘记控制。
输出：基于目前的主线剧情和分线剧情，判断到底要输出什么。
基于上述控制机制，LSTM就延缓了记忆衰退。

可以参考莫凡的LSTM视频讲解，经过实测，稍微改动了一点，对应代码如下：

# View more python learning tutorial on my Youtube and Youku channel!!!

# Youtube video tutorial: https://www.youtube.com/channel/UCdyjiB5H8Pu7aDTNVXTTpcg
# Youku video tutorial: http://i.youku.com/pythontutorial

"""
This code is a modified version of the code from this link:
https://github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/3_NeuralNetworks/recurrent_network.py

His code is a very good one for RNN beginners. Feel free to check it out.
"""
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

# set random seed for comparing the two result calculations
tf.set_random_seed(1)

# this is data
mnist = input_data.read_data_sets('J:\dataset\MNIST_data', one_hot=True)

# hyperparameters
lr = 0.001
training_iters = 100000
batch_size = 128

n_inputs = 28   # MNIST data input (img shape: 28*28)
n_steps = 28    # time steps
n_hidden_units = 128   # neurons in hidden layer
n_classes = 10      # MNIST classes (0-9 digits)

# tf Graph input
x = tf.placeholder(tf.float32, [None, n_steps, n_inputs])
y = tf.placeholder(tf.float32, [None, n_classes])

# Define weights
weights = {
    # (28, 128)
    'in': tf.Variable(tf.random_normal([n_inputs, n_hidden_units])),
    # (128, 10)
    'out': tf.Variable(tf.random_normal([n_hidden_units, n_classes]))
}
biases = {
    # (128, )
    'in': tf.Variable(tf.constant(0.1, shape=[n_hidden_units, ])),
    # (10, )
    'out': tf.Variable(tf.constant(0.1, shape=[n_classes, ]))
}

def RNN(X, weights, biases):
    # hidden layer for input to cell
    ########################################

    # transpose the inputs shape from
    # X ==> (128 batch * 28 steps, 28 inputs)
    X = tf.reshape(X, [-1, n_inputs])

    # into hidden
    # X_in = (128 batch * 28 steps, 128 hidden)
    X_in = tf.matmul(X, weights['in']) + biases['in']
    # X_in ==> (128 batch, 28 steps, 128 hidden)
    X_in = tf.reshape(X_in, [-1, n_steps, n_hidden_units])

    # cell
    ##########################################

    # basic LSTM Cell.
    cell = tf.contrib.rnn.BasicLSTMCell(n_hidden_units)

    # lstm cell is divided into two parts (c_state, h_state)
    init_state = cell.zero_state(batch_size, dtype=tf.float32)

    # You have 2 options for following step.
    # 1: tf.nn.rnn(cell, inputs);
    # 2: tf.nn.dynamic_rnn(cell, inputs).
    # If use option 1, you have to modified the shape of X_in, go and check out this:
    # https://github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/3_NeuralNetworks/recurrent_network.py
    # In here, we go for option 2.
    # dynamic_rnn receive Tensor (batch, steps, inputs) or (steps, batch, inputs) as X_in.
    # Make sure the time_major is changed accordingly.
    outputs, final_state = tf.nn.dynamic_rnn(cell, X_in, initial_state=init_state, time_major=False)

    # hidden layer for output as the final results
    #############################################
    # results = tf.matmul(final_state[1], weights['out']) + biases['out']

    # # or
    # unpack to list [(batch, outputs)..] * steps
    outputs = tf.unstack(tf.transpose(outputs, [1,0,2]))
    results = tf.matmul(outputs[-1], weights['out']) + biases['out']    # shape = (128, 10)

    return results

pred = RNN(x, weights, biases)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pred, labels=y))
train_op = tf.train.AdamOptimizer(lr).minimize(cost)

correct_pred = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))

with tf.Session() as sess:
    # tf.initialize_all_variables() no long valid from
    init = tf.global_variables_initializer()
    sess.run(init)
    step = 0
    while step * batch_size < training_iters:
        batch_xs, batch_ys = mnist.train.next_batch(batch_size)
        batch_xs = batch_xs.reshape([batch_size, n_steps, n_inputs])
        sess.run([train_op], feed_dict={
            x: batch_xs,
            y: batch_ys,
        })
        if step % 20 == 0:
            print(sess.run(accuracy, feed_dict={
            x: batch_xs,
            y: batch_ys,
            }))
        step += 1