I am trying to classify a time series data and after feature extraction the data looks like the following
About this data:
The original data before feature extraction represents accelerometer readings (Image attached in the comments below). What I did was extracting features from each series instead of using the brute data.
In other words, each series/sequence in the original data consisted of 1000 rows so I extracted several features so that I ended up with one row for each series.
I know the new data looks more fit to be used with a Random Forest. However, I was wondering if it was possible to use my new data with a RNN/LSTM/GRU? If yes, how can I do that?
Can you provide some references?
In simpler words, how can I classify some data with several features using RNN in PyTorch?
Thank you very much!