I've trained a neural net to match nx7x5 matrices to nx1 scalars; the n is a number of elements in a sequence. To get a good match I have to do some data prep; basically do a Rescale on each row of the matrix effectivley scaling the data for each row to a number between 0-1. That improves the match quite a lot; but also is a bit tricky and I wanted to use a neural net layer to do that. I'm think BatchNormalizationLayer should allow me to feed the raw data into the network directly? am I correct? I would also like a way to 'see' how the data looks like coming out of the BatchNormalizationLayer to compare it with the input so I can understand it. Below is the link to a small subset of the data.




Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.