Abstract

Traffic accidents rank among the world's most serious concerns due to the high number of deaths, injuries, and fatalities as well as the enormous financial losses they cause every year. Road travel is a necessary component of modern civilization, but because of the rise in traffic accidents, it costs the world economy billions of dollars and over a million deaths annually. Road accidents can be caused by a variety of elements. It can be possible to take action to lessen the severity and extent of the effects if these elements are better recognized and predicted. The goal of this project is to use machine learning techniques to forecast the severity of traffic incidents. Utilizing the US Accidents dataset sample from Kaggle, the project develops a Random Forest classifier prediction model, Decision Tree model, and K-Nearest Neighbor (KNN) model. To predict the severity of accidents, the model utilizes the use of several data, including time-related factors, road characteristics, and weather. Python programming language has been used to develop the predictive model. The project desires to improve public safety and minimize the effect of road traffic events by offering actionable data for emergency response teams and traffic management. The outcomes obtained have increased confidence that the use of advanced features contributes to improved traffic accident prediction. The random forest model has an accuracy of approx. 84%, the Decision Tree model has an accuracy of approx. 76% and K-Nearest Neighbor (KNN) model has an accuracy of approx. 83%. Hence, the random forest model performed well, but there is undoubtedly space for improvement especially when it comes to managing minority classes.

Publication Date

Spring 2024

Document Type

Thesis

Student Type

Graduate

Degree Name

Professional Studies (MS)

Department, Program, or Center

Graduate Programs & Research

Advisor

Sanjay Modak

Campus

RIT Dubai

Download

Share

COinS