Adaptive Conformal Anomaly Detection with Time Series Foundation Models for Signal Monitoring

TL;DR

This paper introduces an adaptive conformal anomaly detection method for time series that uses foundation model predictions, providing interpretable scores and robust performance without additional training.

Contribution

It presents a model-agnostic, post-hoc anomaly detection approach that adaptively calibrates false alarm rates under distribution shifts using weighted conformal bounds.

Findings

Effective in both synthetic and real-world datasets

Provides stable false alarm control under distribution shifts

Seamlessly integrates with foundation models for rapid deployment

Abstract

We propose a post-hoc adaptive conformal anomaly detection method for monitoring time series that leverages predictions from pre-trained foundation models without requiring additional fine-tuning. Our method yields an interpretable anomaly score directly interpretable as a false alarm rate (p-value), facilitating transparent and actionable decision-making. It employs weighted quantile conformal prediction bounds and adaptively learns optimal weighting parameters from past predictions, enabling calibration under distribution shifts and stable false alarm control, while preserving out-of-sample guarantees. As a model-agnostic solution, it integrates seamlessly with foundation models and supports rapid deployment in resource-constrained environments. This approach addresses key industrial challenges such as limited data availability, lack of training expertise, and the need for immediate inference, while taking advantage of the growing accessibility of time series foundation models. Experiments on both synthetic and real-world datasets show that the proposed approach delivers strong performance, combining simplicity, interpretability, robustness, and adaptivity.