search

thuml / iTransformer

Official implementation for "iTransformer: Inverted Transformers Are Effective for Time Series Forecasting" (ICLR 2024 Spotlight), https://openreview.net/forum?id=JePfAI8fah
https://arxiv.org/abs/2310.06625
MIT License
901 stars 168 forks source link

Question: Support for Dynamic Categorical Inputs in iTransformer #61

Open rd1886 opened 3 months ago

rd1886 commented 3 months ago

Hi,

I am interested in using the iTransformer model for a project and had a question about its input capabilities. Does the iTransformer support dynamic categorical inputs? If so, could you provide some guidance or examples on how to implement this? I couldn't find any mention of this feature in the paper.

Thanks!

GuYith commented 3 months ago

I have a similar question, I want to use covariates in training. Does the iTransformer support dynamic covariates? If so, could you provide some guidance or examples on how to implement this?

WenWeiTHU commented 3 months ago

Of course, we are in process now! You may be interested in our recent work where iTransformer is further enhanced with covariates for time series forecasting!

TimeXer: Empowering Transformers for Time Series Forecasting with Exogenous Variables

The code will also be released soon, please stay tuned!