C4.5算法
朴素贝叶斯分类器
支持向量机
交通事故
事故(哲学)
计算机科学
贝叶斯定理
分类器(UML)
道路交通事故
运输工程
训练集
机器学习
人工智能
工程类
贝叶斯概率
哲学
认识论
作者
Tadesse Kebede Bahiru,Manjula V. S,Tadesse Birara Akele,Engdaw Ayalew Tesfaw,Tadesse Destaw Belay
标识
DOI:10.1109/idciot56793.2023.10053409
摘要
As World Health Organization reports show more than 1.3 million people die each year from traffic accidents and more than twenty to fifty million people are harmed by non-fatal accidents. Understanding and identifying the major factors of road traffic accidents can help the stakeholders to take appropriate measures to minimize the accidents. Support Vector Machine (SVM), J48, and Naïve Bayes (NB) classification techniques are implemented to build a model that predicts accident severities and to identify road traffic accident factors. As experimental findings showed weather conditions, number of lanes, road lighting conditions, and speed limit are the determinant factors that cause more accident severity. Finally, the classification and prediction performance of the models are compared using various performance evaluation techniques. J48 classifier has better classification performance than Support Vector Machine and Naïve Bayes classifiers. Naïve Bayes is the classifier that showed the lowest performance than others, but it has good performance in classifying fatal accident severities rather than serious and slight accident severities.
科研通智能强力驱动
Strongly Powered by AbleSci AI