Tutorial: Trustworthy Anomaly Detection

Anomaly detection has a wide range of real-world applications, such as bank fraud detection and cyber intrusion detection. In the past decade, a variety of anomaly detection models have been developed, which lead to big progress towards accurately detecting various anomalies. Despite the successes, anomaly detection models still face many limitations. The most significant one is whether we can trust the detection results from the models. Considering that many anomaly detection tasks are life-changing tasks involving human beings, labeling someone as anomalies or fraudsters should be extremely cautious. Hence, ensuring the anomaly detection models conducted in a trustworthy fashion is an essential requirement to deploy the models to conduct automatic decisions in the real world. In this tutorial, we will introduce the existing efforts and discuss open problems towards trustworthy anomaly detection from the perspectives of explainability, fairness, robustness, and privacy-preservation.

Presenters

• Shuhan Yuan: Assistant Professor in Computer Science Department at Utah State University
• Depeng Xu: Assistant Professor in the Department of Software & Information Systems and the School of Data Science at UNC Charlotte.
• Xintao Wu: Chaired Professor in Computer Science and Computer Engineering Department at University of Arkansas.

Schedule and Materials

1. Introduction
2. Preliminary: Effective Anomaly Detection
3. Explainable Anomaly Detection
4. Fair Anomaly Detection
5. Robust Anomaly Detection
6. Privacy-preservation Anomaly Detection
7. Conclusions and Future Opportunities

Recorded Video

Key References

• Yuan, Shuhan, and Xintao Wu. “Trustworthy anomaly detection: A survey.” arXiv preprint arXiv:2202.07787 (2022).

• Papers on Trustworthy Anomaly Detection