RNN benchmark data reshaping
Closed this issue · 1 comments
shadiakiki1986 commented
Hello. Thanks for sharing this repository.
I'm looking at ML_benchmarks.py#L156 and noticed that you just reshape x_train
to have sequences along each row. This means that your rows do not have any overlap between them. Doesn't this hinder the RNN's performance?
Mcompetitions commented
Hello,
I think the code is fine. L156 is just used for bringing the data-set in the desired format of Keras. The rows are normally overlapped so that the "memory" of the RNN can be exploited during the training. You can test it yourself with an actual t-s of M4.