A generalised label noise model for classification in the presence of annotation errors
© 2016 Elsevier B.V. Supervised learning from annotated data is becoming more challenging due to inherent imperfection of training labels. Previous studies of learning in the presence of label noise have been focused on label noise which occurs randomly, while the study of label noise that is influe...
Saved in:
| Main Author: | |
|---|---|
| Format: | Journal |
| Published: |
2017
|
| Online Access: | https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84959469626&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/41806 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Chiang Mai University |
| id |
th-cmuir.6653943832-41806 |
|---|---|
| record_format |
dspace |
| spelling |
th-cmuir.6653943832-418062017-09-28T04:23:27Z A generalised label noise model for classification in the presence of annotation errors Bootkrajang J. © 2016 Elsevier B.V. Supervised learning from annotated data is becoming more challenging due to inherent imperfection of training labels. Previous studies of learning in the presence of label noise have been focused on label noise which occurs randomly, while the study of label noise that is influenced by input features, which is intuitively more realistic, is still lacking. In this paper, we propose a new, generalised label noise model which is able to withstand the negative effect of random label noise and a wide range of non-random label noises. Empirical studies using a battery of synthetic data and four real-world datasets with inherent annotation errors demonstrate that the proposed generalised label noise model improves, in terms of classification accuracy, upon existing label noise modelling approaches. 2017-09-28T04:23:26Z 2017-09-28T04:23:26Z 2016-06-05 Journal 09252312 2-s2.0-84959469626 10.1016/j.neucom.2015.12.106 https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84959469626&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/41806 |
| institution |
Chiang Mai University |
| building |
Chiang Mai University Library |
| country |
Thailand |
| collection |
CMU Intellectual Repository |
| description |
© 2016 Elsevier B.V. Supervised learning from annotated data is becoming more challenging due to inherent imperfection of training labels. Previous studies of learning in the presence of label noise have been focused on label noise which occurs randomly, while the study of label noise that is influenced by input features, which is intuitively more realistic, is still lacking. In this paper, we propose a new, generalised label noise model which is able to withstand the negative effect of random label noise and a wide range of non-random label noises. Empirical studies using a battery of synthetic data and four real-world datasets with inherent annotation errors demonstrate that the proposed generalised label noise model improves, in terms of classification accuracy, upon existing label noise modelling approaches. |
| format |
Journal |
| author |
Bootkrajang J. |
| spellingShingle |
Bootkrajang J. A generalised label noise model for classification in the presence of annotation errors |
| author_facet |
Bootkrajang J. |
| author_sort |
Bootkrajang J. |
| title |
A generalised label noise model for classification in the presence of annotation errors |
| title_short |
A generalised label noise model for classification in the presence of annotation errors |
| title_full |
A generalised label noise model for classification in the presence of annotation errors |
| title_fullStr |
A generalised label noise model for classification in the presence of annotation errors |
| title_full_unstemmed |
A generalised label noise model for classification in the presence of annotation errors |
| title_sort |
generalised label noise model for classification in the presence of annotation errors |
| publishDate |
2017 |
| url |
https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84959469626&origin=inward http://cmuir.cmu.ac.th/jspui/handle/6653943832/41806 |
| _version_ |
1681422070450225152 |