Lightweight Online Adaption for Time Series Foundation Model Forecasts

TL;DR

This paper introduces ELF, a lightweight online adaptation mechanism for foundation model forecasts in time series, which improves accuracy by effectively utilizing online feedback without significant computational overhead.

Contribution

The paper presents ELF, a novel lightweight method enabling online adaptation of foundation model forecasts using feedback, enhancing performance without high computational costs.

Findings

ELF improves forecast accuracy across multiple datasets.

ELF effectively utilizes online feedback for model adaptation.

Performance gains are consistent with various foundation models.

Abstract

Foundation models (FMs) have emerged as a promising approach for time series forecasting. While effective, FMs typically remain fixed during deployment due to the high computational costs of learning them online. Consequently, deployed FMs fail to adapt their forecasts to current data characteristics, despite the availability of online feedback from newly arriving data. This raises the question of whether FM performance can be enhanced by the efficient usage of this feedback. We propose ELF to answer this question. ELF is a lightweight mechanism for the online adaption of FM forecasts in response to online feedback. ELF consists of two parts: a) the ELF-Forecaster which is used to learn the current data distribution; and b) the ELF-Weighter which is used to combine the forecasts of the FM and the ELF-Forecaster. We evaluate the performance of ELF in conjunction with several recent FMs across a suite of standard time series datasets. In all of our experiments we find that using ELF improves performance. This work demonstrates how efficient usage of online feedback can be used to improve FM forecasts.