A nearest neighbor open-set classifier based on excesses of distance ratios
- Author
- Matthys Lucas Steyn (UGent) , Tertius de Wet, Bernard De Baets (UGent) and Stijn Luca (UGent)
- Organization
- Project
- Abstract
- This article proposes an open-set recognition model that is based on the use of extreme value statistics. For this purpose, a distance ratio is introduced that expresses how dissimilar a target point is from the known classes by considering the ratio of distances locally around the target point. It is shown that the class of generalized Pareto distributions with bounded support can be used to model the peaks of the distance ratio above a high threshold. The resulting distribution provides a probabilistic framework to perform open-set recognition. Furthermore, we describe a numerical procedure to estimate the hyperparameters of our model. This procedure is based on a new objective function that considers both the fit of the generalized Pareto distribution and the misclassification error of the known classes. Our method is applied to three image datasets and an audio dataset showing that it outperforms similar open-set recognition and anomaly detection methods. Supplementary materials for this article are available online.
- Keywords
- Classification, Extreme value theory, Generalized Pareto distribution, Open-set recognition
Downloads
-
A Nearest Neighbor Open set Classifier based on Excesses of Distance Ratios.pdf
- full text (Accepted manuscript)
- |
- open access
- |
- |
- 1.64 MB
-
(...).pdf
- full text (Published version)
- |
- UGent only
- |
- |
- 2.32 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-8762865
- MLA
- Steyn, Matthys Lucas, et al. “A Nearest Neighbor Open-Set Classifier Based on Excesses of Distance Ratios.” JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, vol. 32, no. 1, 2023, pp. 319–28, doi:10.1080/10618600.2022.2096621.
- APA
- Steyn, M. L., de Wet, T., De Baets, B., & Luca, S. (2023). A nearest neighbor open-set classifier based on excesses of distance ratios. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 32(1), 319–328. https://doi.org/10.1080/10618600.2022.2096621
- Chicago author-date
- Steyn, Matthys Lucas, Tertius de Wet, Bernard De Baets, and Stijn Luca. 2023. “A Nearest Neighbor Open-Set Classifier Based on Excesses of Distance Ratios.” JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS 32 (1): 319–28. https://doi.org/10.1080/10618600.2022.2096621.
- Chicago author-date (all authors)
- Steyn, Matthys Lucas, Tertius de Wet, Bernard De Baets, and Stijn Luca. 2023. “A Nearest Neighbor Open-Set Classifier Based on Excesses of Distance Ratios.” JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS 32 (1): 319–328. doi:10.1080/10618600.2022.2096621.
- Vancouver
- 1.Steyn ML, de Wet T, De Baets B, Luca S. A nearest neighbor open-set classifier based on excesses of distance ratios. JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS. 2023;32(1):319–28.
- IEEE
- [1]M. L. Steyn, T. de Wet, B. De Baets, and S. Luca, “A nearest neighbor open-set classifier based on excesses of distance ratios,” JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, vol. 32, no. 1, pp. 319–328, 2023.
@article{8762865,
abstract = {{This article proposes an open-set recognition model that is based on the use of extreme value statistics. For this purpose, a distance ratio is introduced that expresses how dissimilar a target point is from the known classes by considering the ratio of distances locally around the target point. It is shown that the class of generalized Pareto distributions with bounded support can be used to model the peaks of the distance ratio above a high threshold. The resulting distribution provides a probabilistic framework to perform open-set recognition. Furthermore, we describe a numerical procedure to estimate the hyperparameters of our model. This procedure is based on a new objective function that considers both the fit of the generalized Pareto distribution and the misclassification error of the known classes. Our method is applied to three image datasets and an audio dataset showing that it outperforms similar open-set recognition and anomaly detection methods. Supplementary materials for this article are available online.}},
author = {{Steyn, Matthys Lucas and de Wet, Tertius and De Baets, Bernard and Luca, Stijn}},
issn = {{1061-8600}},
journal = {{JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS}},
keywords = {{Classification,Extreme value theory,Generalized Pareto distribution,Open-set recognition}},
language = {{eng}},
number = {{1}},
pages = {{319--328}},
title = {{A nearest neighbor open-set classifier based on excesses of distance ratios}},
url = {{http://doi.org/10.1080/10618600.2022.2096621}},
volume = {{32}},
year = {{2023}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: