JANET: Joint Adaptive predictioN-region Estimation for Time-series

Conformal prediction provides machine learning models with prediction sets that offer theoretical guarantees, but the underlying assumption of exchangeability limits its applicability to time series data. Furthermore, existing approaches struggle to handle multi-step ahead prediction tasks, where un...

Full description

Saved in:
Bibliographic Details
Main Authors English, Eshant, Wong-Toi, Eliot, Fontana, Matteo, Mandt, Stephan, Smyth, Padhraic, Lippert, Christoph
Format Journal Article
LanguageEnglish
Published 08.07.2024
Subjects
Computer Science - Learning
Statistics - Machine Learning
Statistics - Methodology
Online AccessGet full text

Cover

More Information
Summary:Conformal prediction provides machine learning models with prediction sets that offer theoretical guarantees, but the underlying assumption of exchangeability limits its applicability to time series data. Furthermore, existing approaches struggle to handle multi-step ahead prediction tasks, where uncertainty estimates across multiple future time points are crucial. We propose JANET (Joint Adaptive predictioN-region Estimation for Time-series), a novel framework for constructing conformal prediction regions that are valid for both univariate and multivariate time series. JANET generalises the inductive conformal framework and efficiently produces joint prediction regions with controlled K-familywise error rates, enabling flexible adaptation to specific application needs. Our empirical evaluation demonstrates JANET's superior performance in multi-step prediction tasks across diverse time series datasets, highlighting its potential for reliable and interpretable uncertainty quantification in sequential data.
DOI:10.48550/arxiv.2407.06390