Dynamic Multi-period Experts for Online Time Series Forecasting

TL;DR

DynaME (Dynamic Multi-period Experts), a novel hybrid framework designed to effectively address this dual nature of drift, is proposed, which effectively adapts to both concept drifts and significantly outperforms existing baselines.

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.

Figures (5)

• Figure 1: Effect of recurring drift on PROCEED and Weekly-LR performance on the Traffic dataset, showing the time series segment (top) and corresponding MSE (bottom).
• Figure 2: Autocorrelation Function (ACF) analysis on the ECL (top) and Traffic (bottom) datasets. Each plot compares the correlation strength of the daily (Lag-24) and weekly (Lag-168) patterns.
• Figure 3: DynaME architecture. The solid-bordered components ($\phi$, $f_0$, $g$) are parameterized and trained, while the dash-bordered components ($f_1$, $f_2$, $f_3$) are non-parametrically fitted.
• Figure 4: Per-step MSE of "w/$d_t$" and "w/o$d_t$" with danger signal $d_t$ during Emergent Drift. The danger signal rises when an MSE spike is detected.
• Figure 5: Performance comparison with varying Maximum Sample Size ($n$) and Number of Experts ($k$) on the ETTh2 and ETTm1 datasets.