An explainable multi-objective hybrid machine learning model for reducing heart failure mortality.

IF 3.5 4区 计算机科学 Q2 COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE
PeerJ Computer Science Pub Date : 2025-02-25 eCollection Date: 2025-01-01 DOI:10.7717/peerj-cs.2682
F M Javed Mehedi Shamrat, Majdi Khalid, Thamir M Qadah, Majed Farrash, Hanan Alshanbari
{"title":"An explainable multi-objective hybrid machine learning model for reducing heart failure mortality.","authors":"F M Javed Mehedi Shamrat, Majdi Khalid, Thamir M Qadah, Majed Farrash, Hanan Alshanbari","doi":"10.7717/peerj-cs.2682","DOIUrl":null,"url":null,"abstract":"<p><p>As the world grapples with pandemics and increasing stress levels among individuals, heart failure (HF) has emerged as a prominent cause of mortality on a global scale. The most effective approach to improving the chances of individuals' survival is to diagnose this condition at an early stage. Researchers widely utilize supervised feature selection techniques alongside conventional standalone machine learning (ML) algorithms to achieve the goal. However, these approaches may not consistently demonstrate robust performance when applied to data that they have not encountered before, and struggle to discern intricate patterns within the data. Hence, we present a Multi-objective Stacked Enable Hybrid Model (MO-SEHM), that aims to find out the best feature subsets out of numerous different sets, considering multiple objectives. The Stacked Enable Hybrid Model (SEHM) plays the role of classifier and integrates with a multi-objective feature selection method, the Non-dominated Sorting Genetic Algorithm II (NSGA-II). We employed an HF dataset from the Faisalabad Institute of Cardiology (FIOC) and evaluated six ML models, including SEHM with and without NSGA-II for experimental purposes. The Pareto front (PF) demonstrates that our introduced MO-SEHM surpasses the other models, obtaining 94.87% accuracy with the nine relevant features. Finally, we have applied Local Interpretable Model-agnostic Explanations (LIME) with MO-SEHM to explain the reasons for individual outcomes, which makes our model transparent to the patients and stakeholders.</p>","PeriodicalId":54224,"journal":{"name":"PeerJ Computer Science","volume":"11 ","pages":"e2682"},"PeriodicalIF":3.5000,"publicationDate":"2025-02-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11888912/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"PeerJ Computer Science","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.7717/peerj-cs.2682","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/1/1 0:00:00","PubModel":"eCollection","JCR":"Q2","JCRName":"COMPUTER SCIENCE, ARTIFICIAL INTELLIGENCE","Score":null,"Total":0}
引用次数: 0

Abstract

As the world grapples with pandemics and increasing stress levels among individuals, heart failure (HF) has emerged as a prominent cause of mortality on a global scale. The most effective approach to improving the chances of individuals' survival is to diagnose this condition at an early stage. Researchers widely utilize supervised feature selection techniques alongside conventional standalone machine learning (ML) algorithms to achieve the goal. However, these approaches may not consistently demonstrate robust performance when applied to data that they have not encountered before, and struggle to discern intricate patterns within the data. Hence, we present a Multi-objective Stacked Enable Hybrid Model (MO-SEHM), that aims to find out the best feature subsets out of numerous different sets, considering multiple objectives. The Stacked Enable Hybrid Model (SEHM) plays the role of classifier and integrates with a multi-objective feature selection method, the Non-dominated Sorting Genetic Algorithm II (NSGA-II). We employed an HF dataset from the Faisalabad Institute of Cardiology (FIOC) and evaluated six ML models, including SEHM with and without NSGA-II for experimental purposes. The Pareto front (PF) demonstrates that our introduced MO-SEHM surpasses the other models, obtaining 94.87% accuracy with the nine relevant features. Finally, we have applied Local Interpretable Model-agnostic Explanations (LIME) with MO-SEHM to explain the reasons for individual outcomes, which makes our model transparent to the patients and stakeholders.

降低心力衰竭死亡率的可解释多目标混合机器学习模型。
随着世界与流行病和个人压力水平的增加作斗争,心力衰竭(HF)已成为全球范围内死亡的一个主要原因。提高个体生存机会的最有效方法是在早期阶段诊断出这种情况。研究人员广泛使用监督特征选择技术以及传统的独立机器学习(ML)算法来实现这一目标。然而,这些方法在应用于以前没有遇到过的数据时,可能无法始终如一地展示健壮的性能,并且难以识别数据中的复杂模式。因此,我们提出了一个多目标堆叠使能混合模型(MO-SEHM),该模型旨在考虑多个目标,从众多不同的集合中找出最佳特征子集。叠置使能混合模型(Stacked Enable Hybrid Model, SEHM)作为分类器,与多目标特征选择方法non - dominant Sorting Genetic Algorithm II (NSGA-II)相结合。我们使用了费萨拉巴德心脏病研究所(FIOC)的HF数据集,并评估了六种ML模型,包括带有和不带有NSGA-II的SEHM。Pareto front (PF)表明,我们所引入的MO-SEHM模型在9个相关特征上的准确率达到94.87%,优于其他模型。最后,我们将局部可解释模型不可知论解释(LIME)与MO-SEHM应用于解释个体结果的原因,这使得我们的模型对患者和利益相关者透明。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
PeerJ Computer Science
PeerJ Computer Science Computer Science-General Computer Science
CiteScore
6.10
自引率
5.30%
发文量
332
审稿时长
10 weeks
期刊介绍: PeerJ Computer Science is the new open access journal covering all subject areas in computer science, with the backing of a prestigious advisory board and more than 300 academic editors.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信