Authors
Mike Van Ness, Huibin Shen, Hao Wang, Xiaoyong Jin, Danielle C Maddix, Karthick Gopalswamy
Publication date
2023/2/4
Journal
arXiv preprint arXiv:2302.02077
Description
Meta-forecasting is a newly emerging field which combines meta-learning and time series forecasting. The goal of meta-forecasting is to train over a collection of source time series and generalize to new time series one-at-a-time. Previous approaches in meta-forecasting achieve competitive performance, but with the restriction of training a separate model for each sampling frequency. In this work, we investigate meta-forecasting over different sampling frequencies, and introduce a new model, the Continuous Frequency Adapter (CFA), specifically designed to learn frequency-invariant representations. We find that CFA greatly improves performance when generalizing to unseen frequencies, providing a first step towards forecasting over larger multi-frequency datasets.
Total citations
20231
Scholar articles
M Van Ness, H Shen, H Wang, X Jin, DC Maddix… - arXiv preprint arXiv:2302.02077, 2023