Problems in the current research on forecasting with transformers, foundational models, etc.

In this talk, we will discuss some fundamental limitations we perceive in the current operation of foundational models in the context of time series forecasting. We’ll argue that training on ever more data is not always beneficial, and we’ll illustrate how common yet often inadequate evaluation practices can obscure these limitations. Ultimately, we will argue that the way to address these challenges is multimodality.

Slides and some further reading material on my web page:
https://cbergmeir.com/talks/neurips2024/