A generalised label noise model for classification

Learning from labelled data is becoming more and more challenging due to inherent imperfection of training labels. In this paper, we propose a new, generalised label noise model which is able to withstand the negative effect of both random noise and a wide range of non-random label noises. Empirical...

Full description

Saved in:
Bibliographic Details
Main Author: Jakramate Bootkrajang
Format: Conference Proceeding
Published: 2018
Subjects:
Online Access:https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84961807046&origin=inward
http://cmuir.cmu.ac.th/jspui/handle/6653943832/44560
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Chiang Mai University
Description
Summary:Learning from labelled data is becoming more and more challenging due to inherent imperfection of training labels. In this paper, we propose a new, generalised label noise model which is able to withstand the negative effect of both random noise and a wide range of non-random label noises. Empirical studies using three real-world datasets with inherent annotation errors demonstrate that the proposed generalised label noise model improves, in terms of classification accuracy, over existing label noise modelling approaches.