PhD Dissertation Proposal by Sidney TIO Xi Rong | I, Teacher: Augmenting Educational Technologies with Reinforcement Learning

ABOUT THE TALK

Reinforcement Learning (RL) has demonstrated remarkable success in mastering complex domains, from video games to robotic control systems. Following these achievements, researchers have attempted to directly apply RL as teachers in educational contexts, yet these efforts have yielded limited success. We propose that RL's value in education lies not in replacing human teachers, but in augmenting existing educational technologies. This thesis presents a two-pronged approach to integrating RL into educational systems. First, through the framework of Intelligent Tutoring Systems (ITS), we explore how RL agents can serve as surrogate learners, generating meaningful training data to enhance tutoring modules and address real-world implementation challenges. Second, we investigate methods to improve learning-based teaching algorithms, demonstrating how Multi-armed bandits augmented with domain knowledge can deliver more effective learning curricula. Looking forward, we discuss promising research directions, including the integration of Large Language Models to enhance engagement and instructional clarity. This investigation aims to establish practical pathways for leveraging RL's capabilities to enhance, rather than replace, existing educational technologies.

ABOUT THE SPEAKER

Sidney TIO is a Ph.D. candidate in Computer Science at the SMU School of Computing and Information Systems. He is supervised by Professor Pradeep VARAKANTHAM. His research area focused in maximizing training gains for both humans and artificial agents through research in Reinforcement Learning.