TimePred: efficient and interpretable offline change point detection for high volume data -- with application to industrial process monitoring

TL;DR

TimePred is a self-supervised framework that simplifies high-dimensional change-point detection by predicting normalized time indices, enabling efficient, interpretable, and accurate offline detection suitable for industrial applications.

Contribution

It introduces a novel self-supervised approach that reduces multivariate CPD to univariate mean-shift detection, improving efficiency and interpretability in high-volume data.

Findings

Achieves competitive detection performance with significantly reduced computational cost.

Supports feature-level explanations through integrated XAI attribution methods.

Demonstrates practical utility in an industrial manufacturing case study.

Abstract

Change-point detection (CPD) in high-dimensional, large-volume time series is challenging for statistical consistency, scalability, and interpretability. We introduce TimePred, a self-supervised framework that reduces multivariate CPD to univariate mean-shift detection by predicting each sample's normalized time index. This enables efficient offline CPD using existing algorithms and supports the integration of XAI attribution methods for feature-level explanations. Our experiments show competitive CPD performance while reducing computational cost by up to two orders of magnitude. In an industrial manufacturing case study, we demonstrate improved detection accuracy and illustrate the practical value of interpretable change-point insights.