antoinecarme / pyaf

PyAF is an Open Source Python library for Automatic Time Series Forecasting built on top of popular pydata modules.
BSD 3-Clause "New" or "Revised" License
456 stars 73 forks source link

Exogenous data are not used in MLP and LSTM-based models #197

Closed antoinecarme closed 2 years ago

antoinecarme commented 2 years ago

PyAF uses Keras as a backend for including neural network models.

PyAF can build MLP/Dense and LSTM models, but, there is no LSTMX or MLPX model.

Easy to fix.

Target Release : 2022-07-14

antoinecarme commented 2 years ago

PyAF now supports all these models :

       self.mKnownAutoRegressions = ['NoAR' ,
                                      'AR' , 'ARX' ,
                                      'SVR', 'SVRX',
                                      'MLP' , 'MLPX' ,
                                      'LSTM' , 'LSTMX' ,
                                      'XGB' , 'XGBX' ,
                                      'CROSTON', # No CROSTONX for the moment
                                      'LGB', 'LGBX'];
antoinecarme commented 2 years ago

Need to explore 'CROSTONX' feasibility.

antoinecarme commented 2 years ago

FIXED