Saved in:
Bibliographic Details
Main Authors: Ma, Hongwei, Gao, Junbin, Tran, Minh-ngoc
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.17817
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917115233042432
author Ma, Hongwei
Gao, Junbin
Tran, Minh-ngoc
author_facet Ma, Hongwei
Gao, Junbin
Tran, Minh-ngoc
contents Long-horizon multivariate time-series forecasting is challenging because realistic predictions must (i) denoise heterogeneous signals, (ii) track time-varying cross-series dependencies, and (iii) remain stable and physically plausible over long rollout horizons. We present PRISM, which couples a score-based diffusion preconditioner with a dynamic, correlation-thresholded graph encoder and a forecast head regularized by generic physics penalties. We prove contraction of the induced horizon dynamics under mild conditions and derive Lipschitz bounds for graph blocks, explaining the model's robustness. On six standard benchmarks , PRISM achieves consistent SOTA with strong MSE and MAE gains.
format Preprint
id arxiv_https___arxiv_org_abs_2510_17817
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle From Noise to Laws: Regularized Time-Series Forecasting via Denoised Dynamic Graphs
Ma, Hongwei
Gao, Junbin
Tran, Minh-ngoc
Machine Learning
Long-horizon multivariate time-series forecasting is challenging because realistic predictions must (i) denoise heterogeneous signals, (ii) track time-varying cross-series dependencies, and (iii) remain stable and physically plausible over long rollout horizons. We present PRISM, which couples a score-based diffusion preconditioner with a dynamic, correlation-thresholded graph encoder and a forecast head regularized by generic physics penalties. We prove contraction of the induced horizon dynamics under mild conditions and derive Lipschitz bounds for graph blocks, explaining the model's robustness. On six standard benchmarks , PRISM achieves consistent SOTA with strong MSE and MAE gains.
title From Noise to Laws: Regularized Time-Series Forecasting via Denoised Dynamic Graphs
topic Machine Learning
url https://arxiv.org/abs/2510.17817