Saved in:
Bibliographic Details
Main Authors: Zhou, Zhen, Wang, Zhirui, Hong, Qi, Shi, Yunyang, Gu, Ziyuan, Liu, Zhiyuan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.04678
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914305663827968
author Zhou, Zhen
Wang, Zhirui
Hong, Qi
Shi, Yunyang
Gu, Ziyuan
Liu, Zhiyuan
author_facet Zhou, Zhen
Wang, Zhirui
Hong, Qi
Shi, Yunyang
Gu, Ziyuan
Liu, Zhiyuan
contents Time series forecasting in real-world applications requires both high predictive accuracy and interpretable uncertainty quantification. Traditional point prediction methods often fail to capture the inherent uncertainty in time series data, while existing probabilistic approaches struggle to balance computational efficiency with interpretability. We propose a novel Multi-Expert Learning Distributional Labels (LDL) framework that addresses these challenges through mixture-of-experts architectures with distributional learning capabilities. Our approach introduces two complementary methods: (1) Multi-Expert LDL, which employs multiple experts with different learned parameters to capture diverse temporal patterns, and (2) Pattern-Aware LDL-MoE, which explicitly decomposes time series into interpretable components (trend, seasonality, changepoints, volatility) through specialized sub-experts. Both frameworks extend traditional point prediction to distributional learning, enabling rich uncertainty quantification through Maximum Mean Discrepancy (MMD). We evaluate our methods on aggregated sales data derived from the M5 dataset, demonstrating superior performance compared to baseline approaches. The continuous Multi-Expert LDL achieves the best overall performance, while the Pattern-Aware LDL-MoE provides enhanced interpretability through component-wise analysis. Our frameworks successfully balance predictive accuracy with interpretability, making them suitable for real-world forecasting applications where both performance and actionable insights are crucial.
format Preprint
id arxiv_https___arxiv_org_abs_2602_04678
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Let Experts Feel Uncertainty: A Multi-Expert Label Distribution Approach to Probabilistic Time Series Forecasting
Zhou, Zhen
Wang, Zhirui
Hong, Qi
Shi, Yunyang
Gu, Ziyuan
Liu, Zhiyuan
Machine Learning
Artificial Intelligence
Time series forecasting in real-world applications requires both high predictive accuracy and interpretable uncertainty quantification. Traditional point prediction methods often fail to capture the inherent uncertainty in time series data, while existing probabilistic approaches struggle to balance computational efficiency with interpretability. We propose a novel Multi-Expert Learning Distributional Labels (LDL) framework that addresses these challenges through mixture-of-experts architectures with distributional learning capabilities. Our approach introduces two complementary methods: (1) Multi-Expert LDL, which employs multiple experts with different learned parameters to capture diverse temporal patterns, and (2) Pattern-Aware LDL-MoE, which explicitly decomposes time series into interpretable components (trend, seasonality, changepoints, volatility) through specialized sub-experts. Both frameworks extend traditional point prediction to distributional learning, enabling rich uncertainty quantification through Maximum Mean Discrepancy (MMD). We evaluate our methods on aggregated sales data derived from the M5 dataset, demonstrating superior performance compared to baseline approaches. The continuous Multi-Expert LDL achieves the best overall performance, while the Pattern-Aware LDL-MoE provides enhanced interpretability through component-wise analysis. Our frameworks successfully balance predictive accuracy with interpretability, making them suitable for real-world forecasting applications where both performance and actionable insights are crucial.
title Let Experts Feel Uncertainty: A Multi-Expert Label Distribution Approach to Probabilistic Time Series Forecasting
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.04678