Contextual and Seasonal LSTMs for Time Series Anomaly Detection
This work addresses a critical issue in system reliability management for web systems and cloud servers, offering a robust solution for anomaly detection, though it appears incremental as it builds on existing prediction-based methods.
The paper tackled the problem of detecting subtle anomalies like small point anomalies and slowly rising anomalies in univariate time series, which existing methods struggle with, by proposing CS-LSTMs, a prediction-based framework that leverages contextual dependencies and seasonal patterns, achieving state-of-the-art performance on public benchmarks.
Univariate time series (UTS), where each timestamp records a single variable, serve as crucial indicators in web systems and cloud servers. Anomaly detection in UTS plays an essential role in both data mining and system reliability management. However, existing reconstruction-based and prediction-based methods struggle to capture certain subtle anomalies, particularly small point anomalies and slowly rising anomalies. To address these challenges, we propose a novel prediction-based framework named Contextual and Seasonal LSTMs (CS-LSTMs). CS-LSTMs are built upon a noise decomposition strategy and jointly leverage contextual dependencies and seasonal patterns, thereby strengthening the detection of subtle anomalies. By integrating both time-domain and frequency-domain representations, CS-LSTMs achieve more accurate modeling of periodic trends and anomaly localization. Extensive evaluations on public benchmark datasets demonstrate that CS-LSTMs consistently outperform state-of-the-art methods, highlighting their effectiveness and practical value in robust time series anomaly detection.