Framework for Benefit-Based Multiclass Classification

Author:

Sooklal Shellyann1,Hosein Patrick1

Affiliation:

1. University of the West Indies

Abstract

Abstract Health datasets typically comprise of data that are heavily skewed towards the healthy class, thus resulting in classifiers being biased towards this majority class. Due to this imbalance of data, traditional performance metrics, such as accuracy, are not appropriate for evaluating the performance of classifiers with the minority class (disease-affected/unhealthy individuals). In addition, classifiers are trained under the assumption that the costs or benefits associated with different decision outcomes are equal. However, this is usually not the case with health data since it is more important to identify disease affected/unhealthy persons rather than healthy individuals. In this paper we address these problems by examining benefits/costs when evaluating the performance of classifiers. Furthermore, we focus on multiclass classification where the outcome can be one of three or more options. We propose modifications to the Naive Bayes and Logistic Regression algorithms to incorporate costs and benefits for the multiclass scenario as well as compare these to an existing algorithm, hierarchical cost-sensitive kernel logistic regression, and also an adapted hierarchical approach with our cost-benefit based logistic regression model. We demonstrate the effectiveness of all approaches for fetal health classification but the proposed approaches can be applied to any imbalance dataset where benefits and costs are important.

Publisher

Research Square Platform LLC

Reference26 articles.

1. Sooklal, Shellyann and Hosein, Patrick (2020) A Benefit Optimization Approach to the Evaluation of Classification Algorithms. Springer International Publishing, Cham, 978-3-030-36178-5, We address the problem of binary classification when applied to non-communicable diseases. In such problems the data are typically skewed towards samples of healthy subjects. Because of this, traditional performance metrics (such as accuracy) are not suitable. Furthermore, classifiers are typically trained with the assumption that the benefit or cost associated with decision outcomes are the same. In the case of non-communicable diseases this is not necessarily the case since it is more important to err on the side of treatment of the disease rather on the side of over-diagnosis. In this paper we consider the use of benefits/costs for evaluation of classifiers and we also propose how the Logistic Regression cost function can be modified to account for these benefits and costs for better training to achieve the desired goal. We then illustrate the advantage of the approach for the case of identifying diabetes and breast cancer., 35--46, Artificial Intelligence and Applied Mathematics in Engineering Problems, Hemanth, D. Jude and Kose, Utku

2. Xu, Huan (2021) Hierarchical Cost-Sensitive Techniques for Class Imbalance Learning. 10.1109/ICAIBD51990.2021.9459083, 604-609, , , 2021 4th International Conference on Artificial Intelligence and Big Data (ICAIBD)

3. Zhang, Yin and Zhou, Zhi-Hua (2010) Cost-Sensitive Face Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence 32(10): 1758-1769 https://doi.org/10.1109/TPAMI.2009.195

4. Zhu, Ji and Hastie, Trevor (2004) Classification of gene microarrays by penalized logistic regression. Biostatistics 5(3): 427--443 Oxford University Press

5. Machine Learning Mastery. One-vs-Rest and One-vs-One for Multi-Class Classification. https://machinelearningmastery.com/one-vs-rest-and-one-vs-one-for-multi-class-classification/, 2021, april

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3