IMPACT: Influence Modeling for Open-Set Time Series Anomaly Detection
Summary: arXiv:2603.29183v1 Announce Type: cross
Abstract: Open-set anomaly detection (OSAD) is an emerging paradigm designed to utilize limited labeled data from anomaly classes seen in training to identify both seen and unseen anomalies during testing. Current approaches rely on simple augmentation methods to generate pseudo anomalies that replicate unseen anomalies. Despite being promising in image data, these methods are found to be ineffective in time series data due to the failure to preserve its sequential nature, resulting in trivial or unrealistic anomaly patterns. They are further plagued when the training data is contaminated with unlabeled anomalies.
This work introduces IMPACT, a novel framework that leverages influence modeling for open-set time series anomaly detection, to tackle these challenges. The key insight is to:
- i) Learn an influence function that can accurately estimate the impact of individual training samples on the modeling.
- ii) Leverage these influence scores to generate semantically divergent yet realistic unseen anomalies for time series while repurposing high-influential samples as supervised anomalies for anomaly decontamination.
Extensive experiments show that IMPACT significantly outperforms existing state-of-the-art methods, showing superior accuracy under varying OSAD settings and contamination rates. This advancement marks a significant leap in the field of anomaly detection, particularly for time series data, which has unique challenges not present in other data types.
Challenges in Current Anomaly Detection Methods
Traditional anomaly detection methods often struggle with the complexities of time series data. The sequential nature of such data means that patterns can evolve over time, and anomalies are not always static. Current methods, which primarily rely on data augmentation techniques, fail to maintain the integrity of these temporal relationships.
Advantages of the IMPACT Framework
The IMPACT framework offers several advantages over traditional methods:
- Effective Use of Influence Scores: By quantifying the influence of training samples, IMPACT can generate anomalies that are both realistic and relevant to the specific characteristics of the time series data.
- High Accuracy Rates: The framework has been shown to achieve higher accuracy compared to existing approaches, even in scenarios with significant contamination in the training data.
- Adaptability: IMPACT is designed to handle varying conditions of open-set anomaly detection, making it versatile across different applications and datasets.
Conclusion
IMPACT represents a significant step forward in the challenge of open-set anomaly detection for time series data. By focusing on the influence of training samples and creating realistic anomaly patterns, this framework not only enhances detection capabilities but also provides a robust solution to the problems associated with contaminated training datasets. As the field continues to evolve, frameworks like IMPACT will be crucial for advancing anomaly detection methodologies.