Optimum Reject Options for Prototype-based Classification

Fischer L, Hammer B, Wersing H (2015) .

Preprint | English

No fulltext has been uploaded

Author
Abstract
We analyse optimum reject strategies for prototype-based classifiers andreal-valued rejection measures, using the distance of a data point to theclosest prototype or probabilistic counterparts. We compare reject schemes withglobal thresholds, and local thresholds for the Voronoi cells of theclassifier. For the latter, we develop a polynomial-time algorithm to computeoptimum thresholds based on a dynamic programming scheme, and we propose anintuitive linear time, memory efficient approximation thereof with competitiveaccuracy. Evaluating the performance in various benchmarks, we conclude thatlocal reject options are beneficial in particular for simple prototype-basedclassifiers, while the improvement is less pronounced for advanced models. Forthe latter, an accuracy-reject curve which is comparable to support vectormachine classifiers with state of the art reject options can be reached.
Publishing Year
PUB-ID

Cite this

Fischer L, Hammer B, Wersing H. Optimum Reject Options for Prototype-based Classification. 2015.
Fischer, L., Hammer, B., & Wersing, H. (2015). Optimum Reject Options for Prototype-based Classification.
Fischer, L., Hammer, B., and Wersing, H. (2015). Optimum Reject Options for Prototype-based Classification.
Fischer, L., Hammer, B., & Wersing, H., 2015. Optimum Reject Options for Prototype-based Classification.
L. Fischer, B. Hammer, and H. Wersing, “Optimum Reject Options for Prototype-based Classification”, 2015.
Fischer, L., Hammer, B., Wersing, H.: Optimum Reject Options for Prototype-based Classification. (2015).
Fischer, Lydia, Hammer, Barbara, and Wersing, Heiko. “Optimum Reject Options for Prototype-based Classification”. (2015).
This data publication is cited in the following publications:
This publication cites the following data publications:

Export

0 Marked Publications

Open Data PUB

Sources

arXiv 1503.06549

Search this title in

Google Scholar