As a joint project for CS229-Machine Learning and AA228-Decision Making Under Uncertainty I colloborated with my colleagues Ola Shorinwa and Taylor Howell to develop a policy for deploying police cars to track out vehicles with anomalous behavior. Our project used principle component analysis (PCA) to indentify cars within the NGSIM dataset with “anomolous” behavior and then used a Markov Decision Process to decide whether or not to deploy police cars based on a simple cost/reward setup.

Project Report: Learning an Optimal Policy for Police Resource Allocation on Freeways

MATLAB Code: Code is available on GitHub