【RNN以及LSTM的介绍和公式梳理】http://blog.csdn.net/Dark_Scope/article/details/47056361
【知乎 对比 rnn lstm 简单代码] https://www.zhihu.com/question/37082800
[原码]https://gist.github.com/karpathy/d4dee566867f8291f086
【神经网络介绍 sigmoid,relu 区别】https://zhuanlan.zhihu.com/p/21462488?refer=intelligentunit
[多篇分析总结】BP神经网络和RNN神经网络的学习和实现。
[例子学习】人人都能用Python写出LSTM-RNN的代码![你的神经网络学习最佳起步] http://blog.csdn.net/zzukun/article/details/49968129
【LSTM公式推导】http://blog.csdn.net/a635661820/article/details/45390671
【LSTM公式推导2】http://www.cnblogs.com/ooon/p/5594438.html
The unreasonable effective RNN
Google’s Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
Minimal Gated Unit for Recurrent Neural Networks
http://cs231n.github.io/neural-networks-case-study/ [DNN]