Multivariate Time series Anomaly Detection:A Framework of Hidden Markov Models

TL;DR

This paper proposes a framework for detecting anomalies in multivariate time series by transforming them into univariate series using clustering and fuzzy integrals, then applying Hidden Markov Models for detection.

Contribution

It introduces a novel approach combining transformation techniques with HMMs for multivariate anomaly detection, including comparative analysis of different methods.

Findings

Fuzzy C-Means clustering improves transformation quality.

HMM-based detectors effectively identify anomalies.

Comparative analysis highlights the strengths of various transformation methods.

Abstract

In this study, we develop an approach to multivariate time series anomaly detection focused on the transformation of multivariate time series to univariate time series. Several transformation techniques involving Fuzzy C-Means (FCM) clustering and fuzzy integral are studied. In the sequel, a Hidden Markov Model (HMM), one of the commonly encountered statistical methods, is engaged here to detect anomalies in multivariate time series. We construct HMM-based anomaly detectors and in this context compare several transformation methods. A suite of experimental studies along with some comparative analysis is reported.