交通事故死亡率预测的可解释机器学习方法

Md. Asif Khan Rifat, Ahmedul Kabir, Armana Sabiha Huq
{"title":"交通事故死亡率预测的可解释机器学习方法","authors":"Md. Asif Khan Rifat, Ahmedul Kabir, Armana Sabiha Huq","doi":"arxiv-2409.11929","DOIUrl":null,"url":null,"abstract":"Road traffic accidents (RTA) pose a significant public health threat\nworldwide, leading to considerable loss of life and economic burdens. This is\nparticularly acute in developing countries like Bangladesh. Building reliable\nmodels to forecast crash outcomes is crucial for implementing effective\npreventive measures. To aid in developing targeted safety interventions, this\nstudy presents a machine learning-based approach for classifying fatal and\nnon-fatal road accident outcomes using data from the Dhaka metropolitan traffic\ncrash database from 2017 to 2022. Our framework utilizes a range of machine\nlearning classification algorithms, comprising Logistic Regression, Support\nVector Machines, Naive Bayes, Random Forest, Decision Tree, Gradient Boosting,\nLightGBM, and Artificial Neural Network. We prioritize model interpretability\nby employing the SHAP (SHapley Additive exPlanations) method, which elucidates\nthe key factors influencing accident fatality. Our results demonstrate that\nLightGBM outperforms other models, achieving a ROC-AUC score of 0.72. The\nglobal, local, and feature dependency analyses are conducted to acquire deeper\ninsights into the behavior of the model. SHAP analysis reveals that casualty\nclass, time of accident, location, vehicle type, and road type play pivotal\nroles in determining fatality risk. These findings offer valuable insights for\npolicymakers and road safety practitioners in developing countries, enabling\nthe implementation of evidence-based strategies to reduce traffic crash\nfatalities.","PeriodicalId":501301,"journal":{"name":"arXiv - CS - Machine Learning","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"An Explainable Machine Learning Approach to Traffic Accident Fatality Prediction\",\"authors\":\"Md. Asif Khan Rifat, Ahmedul Kabir, Armana Sabiha Huq\",\"doi\":\"arxiv-2409.11929\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Road traffic accidents (RTA) pose a significant public health threat\\nworldwide, leading to considerable loss of life and economic burdens. This is\\nparticularly acute in developing countries like Bangladesh. Building reliable\\nmodels to forecast crash outcomes is crucial for implementing effective\\npreventive measures. To aid in developing targeted safety interventions, this\\nstudy presents a machine learning-based approach for classifying fatal and\\nnon-fatal road accident outcomes using data from the Dhaka metropolitan traffic\\ncrash database from 2017 to 2022. Our framework utilizes a range of machine\\nlearning classification algorithms, comprising Logistic Regression, Support\\nVector Machines, Naive Bayes, Random Forest, Decision Tree, Gradient Boosting,\\nLightGBM, and Artificial Neural Network. We prioritize model interpretability\\nby employing the SHAP (SHapley Additive exPlanations) method, which elucidates\\nthe key factors influencing accident fatality. Our results demonstrate that\\nLightGBM outperforms other models, achieving a ROC-AUC score of 0.72. The\\nglobal, local, and feature dependency analyses are conducted to acquire deeper\\ninsights into the behavior of the model. SHAP analysis reveals that casualty\\nclass, time of accident, location, vehicle type, and road type play pivotal\\nroles in determining fatality risk. These findings offer valuable insights for\\npolicymakers and road safety practitioners in developing countries, enabling\\nthe implementation of evidence-based strategies to reduce traffic crash\\nfatalities.\",\"PeriodicalId\":501301,\"journal\":{\"name\":\"arXiv - CS - Machine Learning\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Machine Learning\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.11929\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Machine Learning","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.11929","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

道路交通事故(RTA)在全球范围内对公共健康构成了严重威胁,造成了巨大的生命损失和经济负担。这在孟加拉国等发展中国家尤为严重。建立可靠的模型来预测交通事故的结果,对于实施有效的预防措施至关重要。为了帮助制定有针对性的安全干预措施,本研究提出了一种基于机器学习的方法,利用达卡大都市交通事故数据库中 2017 年至 2022 年的数据,对致命和非致命道路交通事故结果进行分类。我们的框架采用了一系列机器学习分类算法,包括逻辑回归、支持向量机、奈夫贝叶斯、随机森林、决策树、梯度提升、LightGBM 和人工神经网络。我们采用 SHAP(SHapley Additive exPlanations)方法优先考虑模型的可解释性,该方法阐明了影响事故死亡率的关键因素。结果表明,LightGBM 优于其他模型,其 ROC-AUC 得分为 0.72。我们还进行了全局、局部和特征依赖性分析,以便更深入地了解模型的行为。SHAP 分析表明,伤亡类别、事故时间、地点、车辆类型和道路类型在决定死亡风险方面起着关键作用。这些发现为发展中国家的政策制定者和道路安全从业人员提供了宝贵的见解,有助于实施循证战略,减少交通事故死亡人数。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
An Explainable Machine Learning Approach to Traffic Accident Fatality Prediction
Road traffic accidents (RTA) pose a significant public health threat worldwide, leading to considerable loss of life and economic burdens. This is particularly acute in developing countries like Bangladesh. Building reliable models to forecast crash outcomes is crucial for implementing effective preventive measures. To aid in developing targeted safety interventions, this study presents a machine learning-based approach for classifying fatal and non-fatal road accident outcomes using data from the Dhaka metropolitan traffic crash database from 2017 to 2022. Our framework utilizes a range of machine learning classification algorithms, comprising Logistic Regression, Support Vector Machines, Naive Bayes, Random Forest, Decision Tree, Gradient Boosting, LightGBM, and Artificial Neural Network. We prioritize model interpretability by employing the SHAP (SHapley Additive exPlanations) method, which elucidates the key factors influencing accident fatality. Our results demonstrate that LightGBM outperforms other models, achieving a ROC-AUC score of 0.72. The global, local, and feature dependency analyses are conducted to acquire deeper insights into the behavior of the model. SHAP analysis reveals that casualty class, time of accident, location, vehicle type, and road type play pivotal roles in determining fatality risk. These findings offer valuable insights for policymakers and road safety practitioners in developing countries, enabling the implementation of evidence-based strategies to reduce traffic crash fatalities.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
×
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