Vehicle Accident Foresight System Using Bigdata Intelligent Random Forest Algorithm

Abstract

commercial vehicles are used to improve people life. One of the main concern is the safety of these commercial vehicle during accidents. This research paper presents the idea of a Vehicle Accident Foresight System (VAFS). Accident datasets for VAFS are based on actual vehicle accident datasets extracted from open-source datasets. VAFS aims to analyze past accident data using the random forest algorithm to analyze and predict future accidents. One thousand decision trees are used in implementation. The implemented prototype consists of a dashboard which can specify accident locations according to time, it can also identify the elements which contribute the most to accident severity. The prediction model of VAFS has a minimum accuracy of 69% and a maximum accuracy of 100%.