Pattern-Guided Diffusion Models

ICLR 2026 Conference Submission21607 Authors

Keywords: diffusion, archetypal analysis

TL;DR: This paper presents Pattern-Guided Diffusion Models, which leverage inherent patterns within temporal data to predict future timesteps.

Abstract: Diffusion models have shown promise in forecasting future data from multivariate time series. However, few existing methods account for recurring structures, or patterns, that appear within the data. We present Pattern-Guided Diffusion Models (PGDM), which leverage inherent patterns within temporal data for forecasting future time steps. PGDM first extracts patterns using archetypal analysis and estimates the most likely next pattern in the sequence. By guiding predictions with this pattern estimate, PGDM makes more realistic predictions that fit within the set of known patterns. We additionally introduce a novel uncertainty quantification technique based on archetypal analysis, and we dynamically scale the guidance level based on the pattern estimate uncertainty. We apply our method to two well-motivated forecasting applications, predicting visual field measurements and motion capture frames. On both, we show that pattern guidance reduces PGDM's prediction error by up to 40.67% and 11.10%, respectively. Compared to baselines, PGDM also achieves lower error by up to 65.58% and 82.54%.

Primary Area: learning on time series and dynamical systems

Submission Number: 21607