CAD-DA: Controllable Anomaly Detection after Domain Adaptation by Statistical Inference

Vo Nguyen Le Duy, Hsuan-Tien Lin, Ichiro Takeuchi
Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, PMLR 238:1828-1836, 2024.

Abstract

We propose a novel statistical method for testing the results of anomaly detection (AD) under domain adaptation (DA), which we call CAD-DA—controllable AD under DA. The distinct advantage of the CAD-DA lies in its ability to control the probability of misidentifying anomalies under a pre-specified level $\alpha$ (e.g., 0.05). The challenge within this DA setting is the necessity to account for the influence of DA to ensure the validity of the inference results. We overcome the challenge by leveraging the concept of Selective Inference to handle the impact of DA. To our knowledge, this is the first work capable of conducting a valid statistical inference within the context of DA. We evaluate the performance of the CAD-DA method on both synthetic and real-world datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v238-nguyen-le-duy24a, title = { CAD-DA: Controllable Anomaly Detection after Domain Adaptation by Statistical Inference }, author = {Nguyen Le Duy, Vo and Lin, Hsuan-Tien and Takeuchi, Ichiro}, booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics}, pages = {1828--1836}, year = {2024}, editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen}, volume = {238}, series = {Proceedings of Machine Learning Research}, month = {02--04 May}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v238/nguyen-le-duy24a/nguyen-le-duy24a.pdf}, url = {https://proceedings.mlr.press/v238/nguyen-le-duy24a.html}, abstract = { We propose a novel statistical method for testing the results of anomaly detection (AD) under domain adaptation (DA), which we call CAD-DA—controllable AD under DA. The distinct advantage of the CAD-DA lies in its ability to control the probability of misidentifying anomalies under a pre-specified level $\alpha$ (e.g., 0.05). The challenge within this DA setting is the necessity to account for the influence of DA to ensure the validity of the inference results. We overcome the challenge by leveraging the concept of Selective Inference to handle the impact of DA. To our knowledge, this is the first work capable of conducting a valid statistical inference within the context of DA. We evaluate the performance of the CAD-DA method on both synthetic and real-world datasets. } }
Endnote
%0 Conference Paper %T CAD-DA: Controllable Anomaly Detection after Domain Adaptation by Statistical Inference %A Vo Nguyen Le Duy %A Hsuan-Tien Lin %A Ichiro Takeuchi %B Proceedings of The 27th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2024 %E Sanjoy Dasgupta %E Stephan Mandt %E Yingzhen Li %F pmlr-v238-nguyen-le-duy24a %I PMLR %P 1828--1836 %U https://proceedings.mlr.press/v238/nguyen-le-duy24a.html %V 238 %X We propose a novel statistical method for testing the results of anomaly detection (AD) under domain adaptation (DA), which we call CAD-DA—controllable AD under DA. The distinct advantage of the CAD-DA lies in its ability to control the probability of misidentifying anomalies under a pre-specified level $\alpha$ (e.g., 0.05). The challenge within this DA setting is the necessity to account for the influence of DA to ensure the validity of the inference results. We overcome the challenge by leveraging the concept of Selective Inference to handle the impact of DA. To our knowledge, this is the first work capable of conducting a valid statistical inference within the context of DA. We evaluate the performance of the CAD-DA method on both synthetic and real-world datasets.
APA
Nguyen Le Duy, V., Lin, H. & Takeuchi, I.. (2024). CAD-DA: Controllable Anomaly Detection after Domain Adaptation by Statistical Inference . Proceedings of The 27th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 238:1828-1836 Available from https://proceedings.mlr.press/v238/nguyen-le-duy24a.html.

Related Material