Staff View: Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.

Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.

One of the most studied machine learning challenges that recent studies have shown the susceptibility of deep neural networks to is the class imbalance problem. While concerted research efforts in this direction have been notable in recent years, findings have shown that the canonical learning objec...

Full description

Saved in:

Bibliographic Details
Main Authors:	Mustapha, Ismail, Hasan, Shafaatunnur, Nabbus, Hatem S. Y., Montaser, Mohamed Mostafa Ali, Olatunji, Sunday Olusanya, Shamsuddin, Siti Maryam
Format:	Article
Language:	English
Published:	Science and Information Organization 2023
Subjects:	T Technology (General) T55-55.3 Industrial Safety. Industrial Accident Prevention
Online Access:	http://eprints.utm.my/105382/1/IsmailMustapha2023_InvestigatingGroupDistributionallyRobustOptimization.pdf http://eprints.utm.my/105382/ http://dx.doi.org/10.14569/IJACSA.2023.0140286
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my.utm.105382
record_format	eprints
spelling	my.utm.1053822024-04-24T06:45:54Z http://eprints.utm.my/105382/ Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification. Mustapha, Ismail Hasan, Shafaatunnur Nabbus, Hatem S. Y. Montaser, Mohamed Mostafa Ali Olatunji, Sunday Olusanya Shamsuddin, Siti Maryam T Technology (General) T55-55.3 Industrial Safety. Industrial Accident Prevention One of the most studied machine learning challenges that recent studies have shown the susceptibility of deep neural networks to is the class imbalance problem. While concerted research efforts in this direction have been notable in recent years, findings have shown that the canonical learning objective, empirical risk minimization (ERM), is unable to achieve optimal imbalance learning in deep neural networks given its bias to the majority class. An alternative learning objective, group distributionally robust optimization (gDRO), is investigated in this study for imbalance learning, focusing on tabular imbalanced data as against image data that has dominated deep imbalance learning research. Contrary to minimizing average per instance loss as in ERM, gDRO seeks to minimize the worst group loss over the training data. Experimental findings in comparison with ERM and classical imbalance methods using four popularly used evaluation metrics in imbalance learning across several benchmark imbalance binary tabular data of varying imbalance ratios reveal impressive performance of gDRO, outperforming other compared methods in terms of g-mean and roc-auc. Science and Information Organization 2023 Article PeerReviewed application/pdf en http://eprints.utm.my/105382/1/IsmailMustapha2023_InvestigatingGroupDistributionallyRobustOptimization.pdf Mustapha, Ismail and Hasan, Shafaatunnur and Nabbus, Hatem S. Y. and Montaser, Mohamed Mostafa Ali and Olatunji, Sunday Olusanya and Shamsuddin, Siti Maryam (2023) Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification. International Journal Of Advanced Computer Science And Applications, 14 (2). pp. 739-748. ISSN 2158-107X http://dx.doi.org/10.14569/IJACSA.2023.0140286 DOI: 10.14569/IJACSA.2023.0140286
institution	Universiti Teknologi Malaysia
building	UTM Library
collection	Institutional Repository
continent	Asia
country	Malaysia
content_provider	Universiti Teknologi Malaysia
content_source	UTM Institutional Repository
url_provider	http://eprints.utm.my/
language	English
topic	T Technology (General) T55-55.3 Industrial Safety. Industrial Accident Prevention
spellingShingle	T Technology (General) T55-55.3 Industrial Safety. Industrial Accident Prevention Mustapha, Ismail Hasan, Shafaatunnur Nabbus, Hatem S. Y. Montaser, Mohamed Mostafa Ali Olatunji, Sunday Olusanya Shamsuddin, Siti Maryam Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
description	One of the most studied machine learning challenges that recent studies have shown the susceptibility of deep neural networks to is the class imbalance problem. While concerted research efforts in this direction have been notable in recent years, findings have shown that the canonical learning objective, empirical risk minimization (ERM), is unable to achieve optimal imbalance learning in deep neural networks given its bias to the majority class. An alternative learning objective, group distributionally robust optimization (gDRO), is investigated in this study for imbalance learning, focusing on tabular imbalanced data as against image data that has dominated deep imbalance learning research. Contrary to minimizing average per instance loss as in ERM, gDRO seeks to minimize the worst group loss over the training data. Experimental findings in comparison with ERM and classical imbalance methods using four popularly used evaluation metrics in imbalance learning across several benchmark imbalance binary tabular data of varying imbalance ratios reveal impressive performance of gDRO, outperforming other compared methods in terms of g-mean and roc-auc.
format	Article
author	Mustapha, Ismail Hasan, Shafaatunnur Nabbus, Hatem S. Y. Montaser, Mohamed Mostafa Ali Olatunji, Sunday Olusanya Shamsuddin, Siti Maryam
author_facet	Mustapha, Ismail Hasan, Shafaatunnur Nabbus, Hatem S. Y. Montaser, Mohamed Mostafa Ali Olatunji, Sunday Olusanya Shamsuddin, Siti Maryam
author_sort	Mustapha, Ismail
title	Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
title_short	Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
title_full	Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
title_fullStr	Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
title_full_unstemmed	Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
title_sort	investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.
publisher	Science and Information Organization
publishDate	2023
url	http://eprints.utm.my/105382/1/IsmailMustapha2023_InvestigatingGroupDistributionallyRobustOptimization.pdf http://eprints.utm.my/105382/ http://dx.doi.org/10.14569/IJACSA.2023.0140286
_version_	1797906009156485120
score	13.211869

Investigating group distributionally robust optimization for deep imbalanced learning: a case study of binary tabular data classification.

Similar Items