Dynamic Multi-period Experts for Online Time Series Forecasting

TL;DR

This paper introduces DynaME, a hybrid framework for online time series forecasting that distinguishes between recurring and emergent concept drifts, dynamically adapting models to improve accuracy.

Contribution

It redefines concept drift into two types and proposes a novel hybrid approach that dynamically adapts to each, outperforming existing methods.

Findings

DynaME effectively adapts to both recurring and emergent drifts.

The framework significantly outperforms existing baselines on benchmark datasets.

Dynamic expert selection improves forecasting accuracy.

Abstract

Online Time Series Forecasting (OTSF) requires models to continuously adapt to concept drift. However, existing methods often treat concept drift as a monolithic phenomenon. To address this limitation, we first redefine concept drift by categorizing it into two distinct types: Recurring Drift, where previously seen patterns reappear, and Emergent Drift, where entirely new patterns emerge. We then propose DynaME (Dynamic Multi-period Experts), a novel hybrid framework designed to effectively address this dual nature of drift. For Recurring Drift, DynaME employs a committee of specialized experts that are dynamically fitted to the most relevant historical periodic patterns at each time step. For Emergent Drift, the framework detects high-uncertainty scenarios and shifts reliance to a stable, general expert. Extensive experiments on several benchmark datasets and backbones demonstrate that DynaME effectively adapts to both concept drifts and significantly outperforms existing baselines.