Towards a General Time Series Anomaly Detector with Adaptive Bottlenecks and Dual Adversarial Decoders

2024-05-24

arXiv.org

AI Summary

Existing time-series anomaly detection methods rely on single-dataset training, resulting in poor generalization and limited adaptability to novel domains with scarce labeled data. This paper introduces the first general-purpose, multi-domain anomaly detection model enabling zero-shot cross-domain transfer. Our approach addresses key challenges through three core innovations: (1) an adaptive information bottleneck mechanism for dynamic capacity control; (2) a dual adversarial decoder architecture that explicitly disentangles representations of normal and anomalous patterns; and (3) a multi-domain pretraining framework integrating learnable bottleneck gating, generative adversarial training, and dual-path reconstruction-based discrimination. Evaluated across nine heterogeneous datasets, our model achieves zero-shot performance on par with or surpassing domain-specific supervised models. It significantly enhances robustness and generalization under low-data and cross-domain settings, establishing a new state-of-the-art for universal time-series anomaly detection.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

Time series anomaly detection plays a vital role in a wide range of applications. Existing methods require training one specific model for each dataset, which exhibits limited generalization capability across different target datasets, hindering anomaly detection performance in various scenarios with scarce training data. Aiming at this problem, we propose constructing a general time series anomaly detection model, which is pre-trained on extensive multi-domain datasets and can subsequently apply to a multitude of downstream scenarios. The significant divergence of time series data across different domains presents two primary challenges in building such a general model: (1) meeting the diverse requirements of appropriate information bottlenecks tailored to different datasets in one unified model, and (2) enabling distinguishment between multiple normal and abnormal patterns, both are crucial for effective anomaly detection in various target scenarios. To tackle these two challenges, we propose a General time series anomaly Detector with Adaptive Bottlenecks and Dual Adversarial Decoders (DADA), which enables flexible selection of bottlenecks based on different data and explicitly enhances clear differentiation between normal and abnormal series. We conduct extensive experiments on nine target datasets from different domains. After pre-training on multi-domain data, DADA, serving as a zero-shot anomaly detector for these datasets, still achieves competitive or even superior results compared to those models tailored to each specific dataset.

Problem

• Develops a general time series anomaly detection model.
• Addresses diverse information bottlenecks across datasets.
• Enhances differentiation between normal and abnormal patterns.

Innovation

• Adaptive bottlenecks for diverse datasets
• Dual adversarial decoders for pattern differentiation
• Pre-trained model for zero-shot anomaly detection