Detection of traffic incidents using ensemble SVM and KNN algorithm
Matheswaran, M. Sumathi, B. Al Mrayat, Omar Isam Dutta, Shuvo Mohammed, Obaidur Rahman Singh, Mangal
Matheswaran, M.
Sumathi, B.
Al Mrayat, Omar Isam
Dutta, Shuvo
Mohammed, Obaidur Rahman
Singh, Mangal
Citations
Altmetric:
Other Names
Location
Time Period
Advisors
Original Date
Digitization Date
Issue Date
2025-06-09
Type
Article
Genre
Keywords
Ensemble learning,K-nearest neighbour,Support vector machines,Traffic incidents
Subjects (LCSH)
Citation
M. M, S. B, O. I. Al Mrayat, S. Dutta, O. R. Mohammed and M. Singh, "Detection of Traffic Incidents Using Ensemble SVM and KNN Algorithm," 2025 International Conference on Emerging Engineering Technologies and Applications (IC-EETA), Indore, India, 2025, pp. 671-675, doi: 10.1109/IC-EETA66496.2025.11548064.
Abstract
For intelligent transport cadres, detecting traffic accidents is an important research topic. Many techniques have shown excellent results in identifying traffic accidents. In any case, it is unacceptable to make these techniques stronger. In a specific context, it is not always the case that a method's efficiency improves when applied against another generating index; in fact, it once sounded rather fantastic in a set of data. The suitability of the KNN ensemble and support vector machines (SVM) for traffic incident identification is demonstrated in this work. Due to their excellent neural network performance and ability to generate a nonlinear classifier with the highest generality, SVM and KNN have been proposed as solutions to the traffic incident detection problem. Understanding ensemble learning (EL) is the main goal of this study. The EL approach is being employed to detect traffic accidents. SVM and KNN techniques are used separately at first and then combined to get the intended outcome. For a better final demonstration, a methodology is therefore needed to combine them. The findings of the investigation are shown to show that, out of all the thinking techniques, the best performance is obtained. Better representations of each of its consistency can be found in the suggested approach.
Table of Contents
Description
Click on the DOI link to access this article at the publishers website (may not be free).
Publisher
IEEE