MP-MoE: Matrix Profile-Guided Mixture of Experts for Precipitation Forecasting

2026-03-26

AI Summary

This work proposes the MP-MoE framework to address precipitation biases in numerical weather prediction (NWP) over tropical regions, where complex terrain and convective instability challenge forecast accuracy, and to mitigate the “double penalty” issue inherent in point-wise loss functions when temporal misalignment occurs. By introducing Matrix Profile into precipitation post-processing for the first time, the method constructs a subsequence-level, structure-aware loss function that guides a Mixture-of-Experts (MoE) model to dynamically select experts based on structural similarity. Evaluated on two major river basins in Vietnam, MP-MoE significantly outperforms both raw NWP outputs and existing baselines, achieving higher mean Critical Success Index (CSI-M) for heavy rainfall events and substantially reduced Dynamic Time Warping (DTW) distance, thereby better preserving storm morphology.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Precipitation forecasting remains a persistent challenge in tropical regions like Vietnam, where complex topography and convective instability often limit the accuracy of Numerical Weather Prediction (NWP) models. While data-driven post-processing is widely used to mitigate these biases, most existing frameworks rely on point-wise objective functions, which suffer from the ``double penalty'' effect under minor temporal misalignments. In this work, we propose the Matrix Profile-guided Mixture of Experts (MP-MoE), a framework that integrates conventional intensity loss with a structural-aware Matrix Profile objective. By leveraging subsequence-level similarity rather than point-wise errors, the proposed loss facilitates more reliable expert selection and mitigates excessive penalization caused by phase shifts. We evaluate MP-MoE on rainfall datasets from two major river basins in Vietnam across multiple horizons, including 1-hour intensity and accumulated rainfall over 12, 24, and 48 hours. Experimental results demonstrate that MP-MoE outperforms raw NWP and baseline learning methods in terms of Mean Critical Success Index (CSI-M) for heavy rainfall events, while significantly reducing Dynamic Time Warping (DTW) values. These findings highlight the framework's efficacy in capturing peak rainfall intensities and preserving the morphological integrity of storm events.

Problem

• precipitation forecasting
• numerical weather prediction
• double penalty effect
• tropical regions
• rainfall bias

Innovation

• Matrix Profile
• Mixture of Experts
• precipitation forecasting
• structural-aware loss
• temporal misalignment