ROAD CAR ACCIDENT PREDICTION USING A MACHINE LEARNING ENABLED DATA ANALYSIS

Abstract

Despite fact that vehicle manufacturing has numerous opportunities to design and build shelter structures, road accidents are inevitable. There are numerous accidents in both urban and rural areas. Patterns associated with various situations can be identified by developing reliable prediction models capable of automatically distinguishing distinct unintentional instances. This study aims to develop a robust predictive model for road traffic accidents in India, employing both linear and nonlinear modeling approaches. This paper analyses and predicts road traffic accidents (RTAs) in India using linear, non-linear regression analysis and Multilayer perceptron Neural Networks (MLPNNs), which policymakers can use. It also helps researchers and the field of transportation safety. MLPNN is an advanced technology that has shown great promise in studying ancient data and forecasting forthcoming inclinations. Many predictive models for forecasting the amount of highway accidents have been developed using 26 years of data from 1994 to 2019 for accident counts on India's freeway roads. The best non-linear regression models and MLPNN model was designated for this challenge, with model variables including years and total number of road accidents (in numbers). Performance of the model was contrasted with that of a linear and non-linear regression model developed with the same goal. The findings demonstrate the superiority of the MLPNN model as a forecasting model, with predictions that are reasonably close to forecasting future highway traffic. The study seeks to identify the model that provides superior accuracy and generalization capabilities for road traffic accident prediction, thereby contributing to the development of more effective safety measures and traffic management strategies