Traffic Accidents Severity Prediction using Support Vector Machine Models
Zeinab Farhat1, Ali Karouni2, Bassam Daya3, Pierre Chauvet4, Nizar Hamadeh5
1Zeinab Farhart*, Computer Science, EDST, Lebanese University, Lebanon, France.
2Ali Karouni, Institute University of Technology, Lebanese University, France.
3Bassam Daya, Computer Science, Institute University of Technology, Lebanese University, France.
4Pierre Chauvet, Computer Science, LARIS, Angers University, France.
5Nizar Hmadeh, Computer Science, Institute University of Technology, Lebanese University, France.
Manuscript received on April 20, 2020. | Revised Manuscript received on May 01, 2020. | Manuscript published on May 10, 2020. | PP: 1345-1350 | Volume-9 Issue-7, May 2020. | Retrieval Number: F4393049620/2020©BEIESP | DOI: 10.35940/ijitee.F4393.059720
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In recent years, road traffic accidents (RTA) have become one of the highest national health concerns worldwide. RTA have become the leading cause of losing lives among children and youth. Recent studies have proven that Data Mining Techniques can break down the complexity that prevails between RTA and corresponding factors. In this paper, Support Vector Machine (SVM) based on Radial basis function (RBF) and Linear Kernel Function is applied to predict fatal road accidents in Lebanon. The experimental results reveal that SVM using RBF give the highest accuracy (86%) and the best AUC (86.6%). The obtained decision-making model claims to tackle the fatal RTA phenomenon.
Keywords: Data mining, Prediction, Road Traffic Accidents, SVM.
Scope of the Article: Data mining