Multi scale method improves anomaly detection across time series
No Scale Left Behind: Multi-Scale Autoencoder with Bi-directional Attention for Time Series Anomaly Detection
Machine LearningArtificial Intelligence
Summary
Detecting unusual patterns in data that changes over time is important in many fields like healthcare and finance. These unusual patterns, called anomalies, can happen at very different speeds and lengths of time. The paper introduces a new method called MSCAD that looks at data from multiple time scales at once and lets these scales share information fairly. The authors show that this approach catches anomalies better than many existing methods on a large set of tests.
What this means in practice
- •For industrial monitoring teams: Detect anomalies that occur at different time lengths in sensor data to prevent equipment failures more effectively than previous methods.
- •For financial fraud analysts: Identify unusual transactions and patterns in financial time series data across multiple time scales to improve fraud detection accuracy.
Authors
Jiaheng Guo, Haochen Zhang, Yu-Chao Huang, Jinhao Duan, Nicholas Konz, Tianlong Chen
Abstract
Time series anomaly detection (TSAD) plays a crucial role in healthcare, finance, industrial monitoring, and other sectors. Within and between these settings, anomalies span vastly different temporal scales, from sub-second point spikes to multi-hour drift patterns. However, most existing TSAD methods commit to a single temporal granularity, and multi-scale designs either analyze different scales in isolation or are constrained to a predefined coarse-to-fine hierarchy, both failing to sufficiently capture multi-scale interactions. To resolve this limitation, we propose Multi-Scale Autoencoder with Cross-Scale Attention for TSAD (MSCAD), a simple yet powerful semi-supervised TSAD framework founded on parallel autoencoder branches corresponding to different patch sizes. A stack of symmetric bidirectional cross-scale attention blocks enables every pair of scales to exchange information before reconstruction without allowing any single scale to be privileged. On the comprehensive TSB-AD benchmark (40 datasets, 530 series), MSCAD achieves large performance gains against 50 baselines across multiple metrics, with VUS-PR of 0.57(+9.6%) on the univariate split and 0.47(+9.3%) on the multivariate split compared to the state-of-the-art.