MSc Thesis - Analysis of the Metropolitan Nashville Police Department’s Traffic Stop Practices using R

This is a thesis project completed as part of the ST606 - Msc in Data Science and Analytics- Project & Disseration at National University of Ireland, Maynooth.

Abstract

• Analysed over 3 million traffic stops with 42 features under the supervision of Dr. Niamh Cahill to determine if Racial Profiling exists. Utilized statistical techniques to conclude that African Americans were arrested and searched 16.3% higher compared to other drivers.
• Implemented and explored Machine Learning Regression Models such as Logistic Regression, General Additive Model and Bayesian Hierarchal Model to identify key attributes and investigate comparative results in order to refine best performing model.
• Achieved an accuracy of 95% using Bayesian Hierarchical model, successfully predicting search and arrest trends depending on drivers’ race.

For more information, please refer to the main thesis report.

Author

Raghav Chadha, MSc Data Science and Analytics candidate

Supervisor

Dr. Niamh Cahill , Lecturer / Assistant Professor at Maynooth University

Repository contents

• code/ - source code written for the analysis of the the traffic stop data for the Nashville, Tennesse, US
• report/ - main thesis report, supplementary material and research data management plan
• images/ - various .png files representing the visualisation and modeling results
• data/ - CSV file used for the project
• presentation/ - The presentation was an opportunity to share my work with fellow students and get valuable feedback from staff members and students.