F-SE-LSTM: A Time Series Anomaly Detection Method with Frequency Domain Information

TL;DR

F-SE-LSTM introduces a novel frequency domain approach for time series anomaly detection, leveraging FFT, SENet, and LSTM to improve detection accuracy and efficiency over existing methods.

Contribution

The paper presents a new method combining frequency domain analysis with deep learning for enhanced anomaly detection in time series data.

Findings

Frequency matrix improves anomaly discrimination.

F-SE-LSTM outperforms existing methods in accuracy.

Method demonstrates higher efficiency in detection tasks.

Abstract

With the development of society, time series anomaly detection plays an important role in network and IoT services. However, most existing anomaly detection methods directly analyze time series in the time domain and cannot distinguish some relatively hidden anomaly sequences. We attempt to analyze the impact of frequency on time series from a frequency domain perspective, thus proposing a new time series anomaly detection method called F-SE-LSTM. This method utilizes two sliding windows and fast Fourier transform (FFT) to construct a frequency matrix. Simultaneously, Squeeze-and-Excitation Networks (SENet) and Long Short-Term Memory (LSTM) are employed to extract frequency-related features within and between periods. Through comparative experiments on multiple datasets such as Yahoo Webscope S5 and Numenta Anomaly Benchmark, the results demonstrate that the frequency matrix constructed by F-SE-LSTM exhibits better discriminative ability than ordinary time domain and frequency domain data. Furthermore, F-SE-LSTM outperforms existing state-of-the-art deep learning anomaly detection methods in terms of anomaly detection capability and execution efficiency.