(1 + Epsilon)-Class Classification: An Anomaly Detection Method for Highly Imbalanced or Incomplete Data Sets

Abstract

Anomaly detection is not an easy problem since distribution of anomalous samples is unknown a priori. We explore a novel method that gives a trade-off possibility between one-class and two-class approaches, and leads to a better performance on anomaly detection problems with small or non-representative anomalous samples. The method is evaluated using several data sets and compared to a set of conventional one-class and two-class approaches.

Cite

Text

Borisyak et al. "(1 + Epsilon)-Class Classification: An Anomaly Detection Method for Highly Imbalanced or Incomplete Data Sets." Journal of Machine Learning Research, 2020.

Markdown

[Borisyak et al. "(1 + Epsilon)-Class Classification: An Anomaly Detection Method for Highly Imbalanced or Incomplete Data Sets." Journal of Machine Learning Research, 2020.](https://mlanthology.org/jmlr/2020/borisyak2020jmlr-epsilon/)

BibTeX

@article{borisyak2020jmlr-epsilon,
  title     = {{(1 + Epsilon)-Class Classification: An Anomaly Detection Method for Highly Imbalanced or Incomplete Data Sets}},
  author    = {Borisyak, Maxim and Ryzhikov, Artem and Ustyuzhanin, Andrey and Derkach, Denis and Ratnikov, Fedor and Mineeva, Olga},
  journal   = {Journal of Machine Learning Research},
  year      = {2020},
  pages     = {1-22},
  volume    = {21},
  url       = {https://mlanthology.org/jmlr/2020/borisyak2020jmlr-epsilon/}
}