@article{ref1,
title="Ensemble learning performance analysis on road traffic accident data",
journal="Social Science Research Network electronic library",
year="2022",
author="Seid Yassin, Salahadin",
volume="2022",
number="",
pages="e4160624-e4160624",
abstract="An accident involving a vehicle on the road is one of the most common and deadly accidents that occur every day. In order to minimize the damage, several immediate measures must be taken. Before anything else, it is essential to identify what factors contribute to a serious accident. The first step in improving street safety and the operation of city streets is to find the cause of the accident. We have undertaken this research to investigate an ensemble model to improve the model performance and identify the major contributing factors. As ensemble models, Random Forests, Extra Tree Classifiers, Gradient Boosting, Bagging, and AdaBoost are used to predict road traffic accidents. Single classifiers like Logistic Regression and Decision Trees are also being implemented. As part of the evaluation of the proposed illustrated model, different types of metrics are used to compare each ensemble and single classifier. The experiment showed that Random Forest was capable of beating other ensemble and conventional models in terms of accuracy with 92.7%, precision with 97.2%, recall with 89.64%, and F1 with 93.31%, therefore, proving that to road safety administrations and stakeholders, it is an effective and highly valid method of making reasonable decisions.   Keywords: Classification, predictions, Ensemble Learning, Machine Learning, Random Forest, Road Traffic Accident, Bagging, Boosting<p /> <p>Language: en</p>",
language="en",
issn="1556-5068",
doi="10.2139/ssrn.4160624",
url="http://dx.doi.org/10.2139/ssrn.4160624"
}