Learning kernel logistic regression in the presence of class label noise

The classical machinery of supervised learning machines relies on a correct set of training labels. Unfortunately, there is no guarantee that all of the labels are correct. Labelling errors are increasingly noticeable in today's classification tasks, as the scale and difficulty of these tasks i...

Full description

Saved in:
Bibliographic Details
Main Authors: Jakramate Bootkrajang, Ata Kabán
Format: Journal
Published: 2018
Subjects:
Online Access:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84904348097&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/53422
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
id th-cmuir.6653943832-53422
record_format dspace
spelling th-cmuir.6653943832-534222018-09-04T09:48:58Z Learning kernel logistic regression in the presence of class label noise Jakramate Bootkrajang Ata Kabán Computer Science The classical machinery of supervised learning machines relies on a correct set of training labels. Unfortunately, there is no guarantee that all of the labels are correct. Labelling errors are increasingly noticeable in today's classification tasks, as the scale and difficulty of these tasks increases so much that perfect label assignment becomes nearly impossible. Several algorithms have been proposed to alleviate the problem of which a robust Kernel Fisher Discriminant is a successful example. However, for classification, discriminative models are of primary interest, and rather curiously, the very few existing label-robust discriminative classifiers are limited to linear problems. In this paper, we build on the widely used and successful kernelising technique to introduce a label-noise robust Kernel Logistic Regression classifier. The main difficulty that we need to bypass is how to determine the model complexity parameters when no trusted validation set is available. We propose to adapt the Multiple Kernel Learning approach for this new purpose, together with a Bayesian regularisation scheme. Empirical results on 13 benchmark data sets and two real-world applications demonstrate the success of our approach. © 2014 Elsevier Ltd. 2018-09-04T09:48:58Z 2018-09-04T09:48:58Z 2014-01-01 Journal 00313203 2-s2.0-84904348097 10.1016/j.patcog.2014.05.007 https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84904348097&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/53422
institution Chiang Mai University
building Chiang Mai University Library
country Thailand
collection CMU Intellectual Repository
topic Computer Science
spellingShingle Computer Science
Jakramate Bootkrajang
Ata Kabán
Learning kernel logistic regression in the presence of class label noise
description The classical machinery of supervised learning machines relies on a correct set of training labels. Unfortunately, there is no guarantee that all of the labels are correct. Labelling errors are increasingly noticeable in today's classification tasks, as the scale and difficulty of these tasks increases so much that perfect label assignment becomes nearly impossible. Several algorithms have been proposed to alleviate the problem of which a robust Kernel Fisher Discriminant is a successful example. However, for classification, discriminative models are of primary interest, and rather curiously, the very few existing label-robust discriminative classifiers are limited to linear problems. In this paper, we build on the widely used and successful kernelising technique to introduce a label-noise robust Kernel Logistic Regression classifier. The main difficulty that we need to bypass is how to determine the model complexity parameters when no trusted validation set is available. We propose to adapt the Multiple Kernel Learning approach for this new purpose, together with a Bayesian regularisation scheme. Empirical results on 13 benchmark data sets and two real-world applications demonstrate the success of our approach. © 2014 Elsevier Ltd.
format Journal
author Jakramate Bootkrajang
Ata Kabán
author_facet Jakramate Bootkrajang
Ata Kabán
author_sort Jakramate Bootkrajang
title Learning kernel logistic regression in the presence of class label noise
title_short Learning kernel logistic regression in the presence of class label noise
title_full Learning kernel logistic regression in the presence of class label noise
title_fullStr Learning kernel logistic regression in the presence of class label noise
title_full_unstemmed Learning kernel logistic regression in the presence of class label noise
title_sort learning kernel logistic regression in the presence of class label noise
publishDate 2018
url https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84904348097&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/53422
_version_ 1681424132296671232