Saved in:
Bibliographic Details
Main Authors: Han, Xiao, Zhang, Lu, Wu, Yongkai, Yuan, Shuhan
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2309.16896
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912515290562560
author Han, Xiao
Zhang, Lu
Wu, Yongkai
Yuan, Shuhan
author_facet Han, Xiao
Zhang, Lu
Wu, Yongkai
Yuan, Shuhan
contents Algorithmic recourse provides actionable recommendations to alter unfavorable predictions of machine learning models, enhancing transparency through counterfactual explanations. While significant progress has been made in algorithmic recourse for static data, such as tabular and image data, limited research explores recourse for multivariate time series, particularly for reversing abnormal time series. This paper introduces Recourse in time series Anomaly Detection (RecAD), a framework for addressing anomalies in multivariate time series using backtracking counterfactual reasoning. By modeling the causes of anomalies as external interventions on exogenous variables, RecAD predicts recourse actions to restore normal status as counterfactual explanations, where the recourse function, responsible for generating actions based on observed data, is trained using an end-to-end approach. Experiments on synthetic and real-world datasets demonstrate its effectiveness.
format Preprint
id arxiv_https___arxiv_org_abs_2309_16896
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Algorithmic Recourse in Abnormal Multivariate Time Series
Han, Xiao
Zhang, Lu
Wu, Yongkai
Yuan, Shuhan
Machine Learning
Algorithmic recourse provides actionable recommendations to alter unfavorable predictions of machine learning models, enhancing transparency through counterfactual explanations. While significant progress has been made in algorithmic recourse for static data, such as tabular and image data, limited research explores recourse for multivariate time series, particularly for reversing abnormal time series. This paper introduces Recourse in time series Anomaly Detection (RecAD), a framework for addressing anomalies in multivariate time series using backtracking counterfactual reasoning. By modeling the causes of anomalies as external interventions on exogenous variables, RecAD predicts recourse actions to restore normal status as counterfactual explanations, where the recourse function, responsible for generating actions based on observed data, is trained using an end-to-end approach. Experiments on synthetic and real-world datasets demonstrate its effectiveness.
title Algorithmic Recourse in Abnormal Multivariate Time Series
topic Machine Learning
url https://arxiv.org/abs/2309.16896