Categorical Features for TFT

Nixtla / neuralforecast

Scalable and user friendly neural :brain: forecasting algorithms.

https://nixtlaverse.nixtla.io/neuralforecast

Apache License 2.0

3.05k stars 352 forks source link

Categorical Features for TFT #1098

Open sm-ak-r33 opened 2 months ago

sm-ak-r33 commented 2 months ago

Description

I guess an actual TFT can handle categorical features with embeddings, does nixtla have something similar or it always needs to be done manually with label encodings?

Use case

I am trying to put string in in the input as feature but the model is not being able to parse it.

jmoralez commented 2 months ago

Have you tried providing categorical features (columns of categorical data type)? It wouldn't use an embedding layer but it'd do label encoding.

candalfigomoro commented 2 months ago