New branch for stateful models

[pebasantaf]

Once the current model is completely set, the aim its to applied the necessary changes to update the model as indicated in the following threads:

• D. Moeller. (). Predicting a multiple forward time step of a time series using lstm, [Online]. Available: https://stackoverflow.com/questions/47594861/predicting-a-multiple-forward-time-step- of-a-time-series-using-lstm. (accessed: 2019-12-13).
• (). Stateful lstm and stream predictions, [Online]. Available: https://stackoverflow.com/questions/53190253/stateful-lstm-and-stream-predictions. (accessed: 2019-12-13).
• (). Doubts regarding understanding keras lstms, [Online]. Available: https://stackoverflow.com/questions/53955093/doubts-regarding-understanding-keras-lstms. (accessed: 2019-12-13).

Like this, a "real-time performing" NN would be created hopefully performing as well as the basic model