J T
11/23/2022, 2:31 PMKin Gtz. Olivares
11/23/2022, 2:55 PMn_freq_downsample
parameter.
• In your case with monthly data you can try for example n_freq_downsample=[6,3,1]
.
• It seems that the scale of your data might be challenging too, you might want to try the scaler_type='norm'
to help the optimization of the network (even ReLU nonlinearities might struggle with very large signals).
• You might want to train the network for a bit longer epochs.J T
11/23/2022, 4:11 PMKin Gtz. Olivares
11/23/2022, 4:30 PMfrom neuralforecast.auto import (
AutoMLP, AutoNBEATS,
AutoRNN, AutoTCN, AutoDilatedRNN,
)