Forecast2Anomaly (F2A): Adapting Multivariate Time Series Foundation Models for Anomaly Prediction

TL;DR

Forecast2Anomaly (F2A) enhances pretrained Time Series Foundation Models with a joint forecast-anomaly loss and retrieval-augmented generation to enable accurate, zero-shot anomaly prediction across diverse real-world systems.

Contribution

The paper introduces F2A, a novel framework that adapts TSFMs for anomaly prediction using joint loss fine-tuning and retrieval-augmented prediction, addressing generalization and evolving anomaly challenges.

Findings

F2A outperforms state-of-the-art methods on 16 datasets.

F2A effectively tracks evolving anomalies without model updates.

F2A demonstrates strong zero-shot anomaly prediction capabilities.

Abstract

Forecasting anomalies (anomaly prediction) in multivariate time series from different real-world, dynamic, and complex systems is vital for preempting critical failures, leading to a substantial minimization in operational costs and human labor. Yet, existing methods are limited to specific systems while failing to generalize to evolving anomaly patterns over time. In contrast, pretrained Time Series Foundation Models (TSFMs) have recently demonstrated strong generalization and zero-shot forecasting capabilities. However, their potential remains untapped for anomaly prediction, a task fundamentally different from forecasting normal behavior. Thus, we present Forecast2Anomaly (F2A), a novel framework that empowers TSFMs with anomaly prediction abilities through two key innovations. First, we propose a joint forecast-anomaly loss that fine-tunes TSFMs to accurately forecast future signals even at anomalous time points. Second, we introduce a Retrieval-Augmented Generation (RAG) module that retrieves historically relevant horizons and conditions predictions on them. This component dynamically adapts to distributional shifts at inference time, enabling F2A to track evolving anomalies without requiring model updates. By combining targeted fine-tuning with dynamic retrieval, F2A bridges the gap between robust TSFM zero-shot forecasting and zero-shot anomaly prediction. Extensive experiments across 16 diverse datasets and multiple TSFM backbones show that F2A consistently outperforms state-of-the-art methods, offering a scalable, zero-shot anomaly prediction solution for real-world applications.