Date of Publication : 29th June 2024

Abstract:This research employs a comprehensive approach, utilizing diverse data sources such as historical accident reports, meteorological data, and traffic patterns to proactively predict road accidents through advanced data mining techniques. Employing machine learning methods, including decision trees and neural networks, the study emphasizes meticulous dataset pre-processing to ensure data quality. The proposed predictive model, implemented using a Random Forest classifier, achieves an impressive accuracy of 91.666%. The model integrates real-time and historical data for thorough evaluation, serving as an efficient early warning system for law enforcement and motorists. The research not only contributes to accident prevention but also facilitates policy formulation, traffic management, and urban planning, ultimately enhancing overall road safety and minimizing societal and economic impacts.

Reference :