Within the dynamic field of transportation research, Reinforcement Learning (RL) has been recognized as a critical approach to monitor, model, and manage transportation systems. Among the diverse array of RL techniques, the Upper Confidence Bound (UCB) algorithm stands out for its potential in solving long-standing transportation problems. The UCB method excels in striking a balance between the exploration (trying new things) of new strategies and the exploitation (leveraging known information) of well-estimated rewards, guided by an optimistic methodology toward uncertainty. In this project we will develop UCB methods for Inverse Reinforcement Learning (IRL) and apply them in a variety of Intelligent Transportation System problems in New Jersey. Big data in the form of travel activity records will be purchased and used to build and calibrate the IRL models.