Research on the implementation of LSTM and highway-lstm algorithm (1)

Research on implementation of LSTM and highway-lstm algorithm (1) [Email protected]http://www.cnblogs.com/swje/ Zhouw 2015-12-22

Statement:

1) The LSTM's Learning series is a collection of information from the web of great Daniel and machine learning experts for their selfless dedication. Please refer to the references for specific information. Specific version statements are also referenced in the original literature.

2) This article is for academic exchange only, non-commercial. So each part of the specific reference does not correspond in detail. If a division accidentally violated the interests of everyone, but also look haihan, and contact bloggers deleted.

3) I Caishuxueqian, finishing summary of the time is inevitable error, but also hope that the predecessors, thank you.

Please contact: [email protected] or [email protected]

The subject of this research is my undergraduate graduation design, open the title Report preliminary determination, the back will be updated in real-time, hoping to communicate with you and common progress! first, research projects:1. Understand the basic concepts, principles and applications of LSTM 2, analyze Lstm code Implementation 3, understand the highway of deep learning, and how to implement HIGHWAY4 in Lstm, analyze highway-lstm code 5, select an application to verify Regular (every January) to summarize the progress of the algorithm analysis, the target at the end of March, fully grasp the LSTM code and algorithm. Second, the following is part Learning MaterialsThe link: Resources on GitHub:Basic LSTM (LUA) https://github.com/wojzaremba/lstm basic LSTM by Torch (LUA) https://github.com/karpathy/char-rnn HIGHW Ay LSTM by Torch (LUA) https://github.com/yoonkim/lstm-char-cnn lstm:http://colah.github.io/posts/2015-08- Understanding-lstms/lstm implementation:http://apaszke.github.io/posts.html

Some references:

[High-way-lstm-1] High-way LSTM Rnns for distant Speech recognition.pdf: http://yunpan.cn/cuAYRBGLtX6UJ access password 510b

[High-way-lstm-2] Character-aware Neural Language models.pdf:http://yunpan.cn/cuAYRBGLtX6UJ access password 510b

Research on the implementation of LSTM and highway-lstm algorithm (1)