Road Accident Analysis Using Machine Learning

Abstract

Road accidents are one among the key concerns within the country. The economic impact of traffic accidents cost many billions of dollars for US citizens per annum. Reducing the traffic accidents has always been a challenge. The most important objective of the project is to investigate what are the foundations for mishaps like effect of precipitation or any natural elements and to predict the probability of road accidents supported the present accidents records. By performing data visualization, we are able to identify the car accidents zones, factors effecting accident severity, and locations etc. Identifying the key factors of how these accidents occur can help in implementing well informed actions. Road safety analysts following up on street accidents information have seen outcome in street auto collisions examination through the applying information logical methods, however, little headway was the expectation of street injury. This report says progressed information investigation strategies to anticipate injury seriousness levels and assesses their presentation. The review utilizes prescient demonstrating methods to recognize chance and key factors that adds to mishap seriousness. The review utilizes openly accessible information from US branch of transport that covers the sum from 2005 to 2019. The report presents a methodology which is adequately general so are frequently applied to various informational collections from different nations. The outcomes recognized that tree based methods like Extreme Gradient Boosting beat relapse covers, as Artificial Neural Network (ANN) also to the paper, recognizes fascinating connections were recognized concepts related with nature of information.