FreCT: Frequency-augmented Convolutional Transformer for Robust Time Series Anomaly Detection

TL;DR

FreCT is a novel model that combines frequency analysis with a convolutional transformer to improve the robustness and accuracy of time series anomaly detection across various domains.

Contribution

The paper introduces FreCT, a frequency-augmented convolutional transformer that integrates Fourier analysis and contrastive learning to enhance anomaly detection in time series data.

Findings

FreCT outperforms existing methods on four public datasets.

The frequency analysis improves the model's ability to capture essential features.

Contrastive views and robust training enhance detection accuracy.

Abstract

Time series anomaly detection is critical for system monitoring and risk identification, across various domains, such as finance and healthcare. However, for most reconstruction-based approaches, detecting anomalies remains a challenge due to the complexity of sequential patterns in time series data. On the one hand, reconstruction-based techniques are susceptible to computational deviation stemming from anomalies, which can lead to impure representations of normal sequence patterns. On the other hand, they often focus on the time-domain dependencies of time series, while ignoring the alignment of frequency information beyond the time domain. To address these challenges, we propose a novel Frequency-augmented Convolutional Transformer (FreCT). FreCT utilizes patch operations to generate contrastive views and employs an improved Transformer architecture integrated with a convolution module to capture long-term dependencies while preserving local topology information. The introduced frequency analysis based on Fourier transformation could enhance the model's ability to capture crucial characteristics beyond the time domain. To protect the training quality from anomalies and improve the robustness, FreCT deploys stop-gradient Kullback-Leibler (KL) divergence and absolute error to optimize consistency information in both time and frequency domains. Extensive experiments on four public datasets demonstrate that FreCT outperforms existing methods in identifying anomalies.