batch_size_input_data

[EPinzuti]

Dear Jules,
thanks a lot for this work. I have one question. I am dealing with data with 3 dimensions, trials x time x features. The trials are repetitions of the same underlying "system". To make use of that, I was thinking to reshape my input data as (trials x time) x features and use a batch size like batch=time. In this way multiple trials are passed to the autoencoder. Does it make sense ? thanks again