KDSelector: A Knowledge-Enhanced and Data-Efficient Model Selector Learning Framework for Time Series Anomaly Detection
Zhiyu Liang, Dongrui Cai, Chenyuan Zhang, Zheng Liang, Chen Liang, Bo Zheng, Shi Qiu, Jin Wang, Hongzhi Wang
TL;DR
KDSelector tackles the TSAD model selection problem by enriching NN-based selectors with knowledge-augmented learning and data-efficient training. It introduces three plug-and-play components—PISL for soft, performance-informed targets; MKI for integrating diverse metadata through mutual information with a language-model representation; and PA for pruning redundant samples via bucketed sampling and gradient rescaling. Together, these modules improve selection accuracy and reduce training time while remaining architecture-agnostic. The framework is demonstrated in an end-to-end system with 12 TSAD models and 16 datasets, showing superior model selection performance and faster training, making it practical for real-world, heterogeneous time series data.
Abstract
Model selection has been raised as an essential problem in the area of time series anomaly detection (TSAD), because there is no single best TSAD model for the highly heterogeneous time series in real-world applications. However, despite the success of existing model selection solutions that train a classification model (especially neural network, NN) using historical data as a selector to predict the correct TSAD model for each series, the NN-based selector learning methods used by existing solutions do not make full use of the knowledge in the historical data and require iterating over all training samples, which limits the accuracy and training speed of the selector. To address these limitations, we propose KDSelector, a novel knowledge-enhanced and data-efficient framework for learning the NN-based TSAD model selector, of which three key components are specifically designed to integrate available knowledge into the selector and dynamically prune less important and redundant samples during the learning. We develop a TSAD model selection system with KDSelector as the internal, to demonstrate how users improve the accuracy and training speed of their selectors by using KDSelector as a plug-and-play module. Our demonstration video is hosted at https://youtu.be/2uqupDWvTF0.