A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning

© 2019 Elsevier Inc. The evidential K nearest neighbor classifier is based on discounting evidence from learning instances in a neighborhood of the pattern to be classified. To adapt the method to partially supervised data, we propose to replace the classical discounting operation by contextual disc...

Full description

Saved in:
Bibliographic Details
Main Authors: Thierry Denœux, Orakanya Kanjanatarakul, Songsak Sriboonchitta
Format: Journal
Published: 2020
Subjects:
Online Access:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85073707748&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/67705
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
id th-cmuir.6653943832-67705
record_format dspace
spelling th-cmuir.6653943832-677052020-04-02T15:10:36Z A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning Thierry Denœux Orakanya Kanjanatarakul Songsak Sriboonchitta Computer Science Mathematics © 2019 Elsevier Inc. The evidential K nearest neighbor classifier is based on discounting evidence from learning instances in a neighborhood of the pattern to be classified. To adapt the method to partially supervised data, we propose to replace the classical discounting operation by contextual discounting, a more complex operation based on as many discount rates as classes. The parameters of the method are tuned by maximizing the evidential likelihood, an extension of the likelihood function based on uncertain data. The resulting classifier is shown to outperform alternative methods in partially supervised learning tasks. 2020-04-02T15:01:44Z 2020-04-02T15:01:44Z 2019-10-01 Journal 0888613X 2-s2.0-85073707748 10.1016/j.ijar.2019.07.009 https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85073707748&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/67705
institution Chiang Mai University
building Chiang Mai University Library
country Thailand
collection CMU Intellectual Repository
topic Computer Science
Mathematics
spellingShingle Computer Science
Mathematics
Thierry Denœux
Orakanya Kanjanatarakul
Songsak Sriboonchitta
A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
description © 2019 Elsevier Inc. The evidential K nearest neighbor classifier is based on discounting evidence from learning instances in a neighborhood of the pattern to be classified. To adapt the method to partially supervised data, we propose to replace the classical discounting operation by contextual discounting, a more complex operation based on as many discount rates as classes. The parameters of the method are tuned by maximizing the evidential likelihood, an extension of the likelihood function based on uncertain data. The resulting classifier is shown to outperform alternative methods in partially supervised learning tasks.
format Journal
author Thierry Denœux
Orakanya Kanjanatarakul
Songsak Sriboonchitta
author_facet Thierry Denœux
Orakanya Kanjanatarakul
Songsak Sriboonchitta
author_sort Thierry Denœux
title A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
title_short A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
title_full A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
title_fullStr A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
title_full_unstemmed A new evidential K-nearest neighbor rule based on contextual discounting with partially supervised learning
title_sort new evidential k-nearest neighbor rule based on contextual discounting with partially supervised learning
publishDate 2020
url https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85073707748&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/67705
_version_ 1681426684650192896