Multi-scale Anomaly Detection for Big Time Series of Industrial Sensors

TL;DR

This paper introduces MissGAN, a multi-scale reconstruction-based anomaly detection method for large multivariate time series, effectively capturing semantic segments without arbitrary cuts, and demonstrating superior performance on industrial sensor data.

Contribution

The paper proposes MissGAN, a novel multi-scale anomaly detection framework that learns from coarse and fine segments using adversarial regularization and HMM, applicable with minimal labeling.

Findings

Outperforms baseline methods on industrial sensor datasets

Effectively detects anomalies without arbitrary segmentation

Successfully distinguishes unexpected gestures in motion data

Abstract

Given a multivariate big time series, can we detect anomalies as soon as they occur? Many existing works detect anomalies by learning how much a time series deviates away from what it should be in the reconstruction framework. However, most models have to cut the big time series into small pieces empirically since optimization algorithms cannot afford such a long series. The question is raised: do such cuts pollute the inherent semantic segments, like incorrect punctuation in sentences? Therefore, we propose a reconstruction-based anomaly detection method, MissGAN, iteratively learning to decode and encode naturally smooth time series in coarse segments, and finding out a finer segment from low-dimensional representations based on HMM. As a result, learning from multi-scale segments, MissGAN can reconstruct a meaningful and robust time series, with the help of adversarial regularization and extra conditional states. MissGAN does not need labels or only needs labels of normal instances, making it widely applicable. Experiments on industrial datasets of real water network sensors show our MissGAN outperforms the baselines with scalability. Besides, we use a case study on the CMU Motion dataset to demonstrate that our model can well distinguish unexpected gestures from a given conditional motion.