Awesome
Description<br>
There are my open-souce code by Keras according to the paper <br>
XinLi,LidongBing,WaiLam and BeiShi. A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction
<br>
layer_definition<br>
This part includes writing own layers.<br>
dual_attention<br>
This part includes data's preprocessing, model's building and model's training.<br>
data
Downloaded from (https://cseweb.ucsd.edu/~yaq007/NASDAQ100_stock_data.html)