NeuCoReClass AD: Redefining Self-Supervised Time Series Anomaly Detection

2025-07-29

AI Summary

Existing self-supervised time-series anomaly detection methods predominantly rely on a single pretext task and domain-specific handcrafted transformations, resulting in limited pattern modeling capability and poor cross-domain generalizability. To address these limitations, we propose TS-MultiSSL, a multi-task self-supervised framework that jointly optimizes three complementary pretext tasks: contrastive learning, sequence reconstruction, and anomaly-type classification. Crucially, TS-MultiSSL incorporates neural transformation learning to automatically generate diverse yet semantically consistent augmented views—eliminating the need for domain-specific prior knowledge. The end-to-end model holistically captures normal temporal patterns and enables fine-grained anomaly-type discrimination in a fully unsupervised setting. Extensive experiments on multiple benchmark datasets demonstrate that TS-MultiSSL achieves state-of-the-art detection performance, while exhibiting superior generalizability across domains and enhanced interpretability through its multi-task design and learned transformations.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Multi-modal Vision
• Multiagent Systems: Multiagent Learning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Time series anomaly detection plays a critical role in a wide range of real-world applications. Among unsupervised approaches, self-supervised learning has gained traction for modeling normal behavior without the need of labeled data. However, many existing methods rely on a single proxy task, limiting their ability to capture meaningful patterns in normal data. Moreover, they often depend on handcrafted transformations tailored specific domains, hindering their generalization accross diverse problems. To address these limitations, we introduce NeuCoReClass AD, a self-supervised multi-task time series anomaly detection framework that combines contrastive, reconstruction, and classification proxy tasks. Our method employs neural transformation learning to generate augmented views that are informative, diverse, and coherent, without requiring domain-specific knowledge. We evaluate NeuCoReClass AD across a wide range of benchmarks, demonstrating that it consistently outperforms both classical baselines and most deep-learning alternatives. Furthermore, it enables the characterization of distinct anomaly profiles in a fully unsupervised manner.

Problem

• Overcoming single proxy task limitations in anomaly detection
• Eliminating need for domain-specific handcrafted transformations
• Enhancing unsupervised anomaly profiling across diverse datasets

Innovation

• Combines contrastive, reconstruction, and classification tasks
• Uses neural transformation learning for diverse augmentations
• Detects anomalies without domain-specific knowledge