Static and dynamic feature for hybrid modle
I would like to make a hybrid model in this way :

I have a question about input data.
My dynamic input is like [ [ [5],[4],[7] ],[ [7],[5],[7] ],...]. So shape is (400,3,1).
But I don't know how the shape should be for my static input. I'm thinking something like this [ [1],[2],...], shape is (400,1). So for this sequence [ [5],[4],[7] ] in LSTM, i have this [1] in static_input.
Do you think it's good?
Topic lstm tensorflow python
Category Data Science