Papers
Topics
Authors
Recent
2000 character limit reached

About Test-time training for outlier detection (2404.03495v1)

Published 4 Apr 2024 in cs.LG

Abstract: In this paper, we introduce DOUST, our method applying test-time training for outlier detection, significantly improving the detection performance. After thoroughly evaluating our algorithm on common benchmark datasets, we discuss a common problem and show that it disappears with a large enough test set. Thus, we conclude that under reasonable conditions, our algorithm can reach almost supervised performance even when no labeled outliers are given.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (51)
  1. On model performance estimation in time series anomaly detection. In 2023 6th International Conference on Computational Intelligence and Intelligent Systems, 2023.
  2. Learning from positive and unlabeled data: a survey. Machine Learning, 109(4):719–760, Apr 2020.
  3. One-class support vector machines revisited. In 2014 International Conference on Information Science Applications (ICISA), pages 1–4, 2014.
  4. Area under the precision-recall curve: Point estimates and confidence intervals. In Hendrik Blockeel, Kristian Kersting, Siegfried Nijssen, and Filip Železný, editors, Machine Learning and Knowledge Discovery in Databases, pages 451–466, Berlin, Heidelberg, 2013. Springer Berlin Heidelberg.
  5. Leo Breiman. Random forests. Machine learning, 45, 2001.
  6. Lof: Identifying density-based local outliers. volume 29, pages 93–104, 06 2000.
  7. Post-robustifying deep anomaly detection ensembles by model selection. In ICDM, 2022.
  8. Event-based anomaly detection for new physics searches at the LHC using machine learning. In APS April Meeting Abstracts, volume 2022 of APS Meeting Abstracts, page Q09.001, January 2022.
  9. Outlier Detection with Autoencoder Ensembles, pages 90–98. 06 2017.
  10. A method of anomaly detection and fault diagnosis with online adaptive learning under small training samples. Pattern Recognition, 64, 2017.
  11. Milton Friedman. A comparison of alternative tests of significance for the problem of $m$ rankings. Annals of Mathematical Statistics, 11:86–92, 1940.
  12. Test-time training with masked autoencoders. In Alice H. Oh, Alekh Agarwal, Danielle Belgrave, and Kyunghyun Cho, editors, Advances in Neural Information Processing Systems, 2022.
  13. A probabilistic interpretation of precision, recall and f-score, with implication for evaluation. volume 3408, pages 345–359, 04 2005.
  14. Statistical analysis of nearest neighbor methods for anomaly detection. In Hanna M. Wallach, Hugo Larochelle, Alina Beygelzimer, Florence d’Alché Buc, Emily B. Fox, and Roman Garnett, editors, NeurIPS, pages 10921–10931, 2019.
  15. Rajib Kumar Halder. Cardiovascular disease dataset, 2020.
  16. Adbench: Anomaly detection benchmark. In NeurIPS, 2022.
  17. The meaning and use of the area under a receiver operating characteristic (roc) curve. Radiology, 143:29–36, 05 1982.
  18. N.R. Hanson. The Concept of the Positron: A Philosophical Analysis. Cambridge University Press, 1963.
  19. Discovering cluster-based local outliers. Pattern Recognition Letters, 24(9):1641–1650, 2003.
  20. The extent and consequences of p-hacking in science. PLoS Biology, 13, 2015.
  21. Financial fraud: a review of anomaly detection techniques and recent advances. Expert systems With applications, 193, 2022.
  22. Sture Holm. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6(2):65–70, 1979.
  23. A survey of recent trends in one class classification. In Lorcan Coyle and Jill Freyne, editors, Artificial Intelligence and Cognitive Science, pages 188–197, Berlin, Heidelberg, 2010. Springer Berlin Heidelberg.
  24. Evaluating and comparing heterogeneous ensemble methods for unsupervised anomaly detection. In IJCNN, 2023.
  25. Feature bagging for outlier detection. volume 21, pages 157–166, 01 2005.
  26. Learning to identify unexpected instances in the test set. In International Joint Conference on Artificial Intelligence, 2007.
  27. A comprehensive survey on test-time adaptation under distribution shifts. ArXiv, abs/2303.15361, 2023.
  28. Isolation forest. In ICDM, 2008.
  29. Deep unsupervised domain adaptation: A review of recent advances and perspectives. APSIPA Transactions on Signal and Information Processing, 05 2022.
  30. The need for unsupervised outlier model selection: A review and evaluation of internal evaluation strategies. ACM SIGKDD Explorations Newsletter, 25(1), 2023.
  31. Viktor Malyi. Run or walk dataset, Jul 2017.
  32. A survey on gans for anomaly detection, 2021.
  33. Online-compatible unsupervised nonresonant anomaly detection. Phys. Rev. D, 105:055006, Mar 2022.
  34. Dubravko Miljković. Fault detection methods: A literature survey. In 2011 Proceedings of the 34th International Convention MIPRO, pages 750–755, 2011.
  35. Anomaly detection in batch chemical processes. In Jacek Jeżowski and Jan Thullie, editors, 19th European Symposium on Computer Aided Process Engineering, volume 26 of Computer Aided Chemical Engineering, pages 255–260. Elsevier, 2009.
  36. Positive unlabeled learning-based anomaly detection in videos. International Journal of Intelligent Systems, 36, 05 2021.
  37. Meta-survey on outlier and anomaly detection. Neurocomputing, 555:126634, 2023.
  38. Scikit-learn: Machine learning in python. Journal of machine learning research, 12(Oct):2825–2830, 2011.
  39. Latent outlier exposure for anomaly detection with contaminated data. In International Conference on Machine Learning, 2022.
  40. Cross-Validation, pages 532–538. Springer US, Boston, MA, 2009.
  41. A unifying review of deep and shallow anomaly detection. Proceedings of the IEEE, PP:1–40, 02 2021.
  42. Deep one-class classification. In ICML, 2018.
  43. Deep semi-supervised anomaly detection. CoRR, abs/1906.02694, 2019.
  44. Anomaly detection using autoencoders with nonlinear dimensionality reduction. In Proceedings of the MLSDA 2014 2nd Workshop on Machine Learning for Sensory Data Analysis, MLSDA’14, page 4–11, New York, NY, USA, 2014. Association for Computing Machinery.
  45. Test-time training with self-supervision for generalization under distribution shifts. In Proceedings of the 37th International Conference on Machine Learning, ICML’20. JMLR.org, 2020.
  46. Learning competitive and discriminative reconstructions for anomaly detection. In AAAI Conference on Artificial Intelligence, 2019.
  47. Frank Wilcoxon. Individual comparisons by ranking methods. Biometrics Bulletin, 1(6):80–83, 1945.
  48. A comparative survey of deep active learning. ArXiv, abs/2203.13450, 2022.
  49. Positive and unlabeled learning for anomaly detection with multi-features. pages 854–862, 10 2017.
  50. Xgbod: Improving supervised outlier detection with unsupervised representation learning. In 2018 International Joint Conference on Neural Networks (IJCNN). IEEE, July 2018.
  51. Pyod: A python toolbox for scalable outlier detection. Journal of Machine Learning Research, 20(96):1–7, 2019.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
Dice Question Streamline Icon: https://streamlinehq.com

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now
Lightbulb Streamline Icon: https://streamlinehq.com

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now
List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets

This paper has been mentioned in 1 tweet and received 0 likes.

Upgrade to Pro to view all of the tweets about this paper:

Start a free 7-day Pro trial