ECATS: Explainable-by-design concept-based anomaly detection for time series

TL;DR

ECATS is a neuro-symbolic, concept-based anomaly detection model for time series that combines STL formulae with kernel methods to provide interpretable predictions in Cyber Physical Systems.

Contribution

It introduces ECATS, a novel architecture that integrates STL-based concepts with neural networks for explainable anomaly detection in time series.

Findings

Achieves high classification performance on CPS data

Provides meaningful local explanations for predictions

Combines unsupervised concept learning with neural classification

Abstract

Deep learning methods for time series have already reached excellent performances in both prediction and classification tasks, including anomaly detection. However, the complexity inherent in Cyber Physical Systems (CPS) creates a challenge when it comes to explainability methods. To overcome this inherent lack of interpretability, we propose ECATS, a concept-based neuro-symbolic architecture where concepts are represented as Signal Temporal Logic (STL) formulae. Leveraging kernel-based methods for STL, concept embeddings are learnt in an unsupervised manner through a cross-attention mechanism. The network makes class predictions through these concept embeddings, allowing for a meaningful explanation to be naturally extracted for each input. Our preliminary experiments with a simple CPS-based dataset show that our model is able to achieve great classification performance while ensuring local interpretability.