Jaehyun Lim

Papers

Total Citations

H-Index

About

Jaehyun Lim is a researcher whose work sits at the intersection of reinforcement learning, control theory, and robotics, with a particular focus on leveraging Gaussian processes (GPs) for intelligent decision-making under uncertainty. His most impactful contribution is the paper "Prediction of Reward Functions for Deep Reinforcement Learning via Gaussian Process Regression" (2020, 46 citations), which tackles the challenge of inverse reinforcement learning (IRL) in high-dimensional environments with unknown dynamics. By proposing a sparse Gaussian process prediction method with l1 regularization, Lim offers an efficient pathway for automatic reward acquisition—a critical step for training agents in complex, real-world tasks. His work also extends to control, as seen in "Nonlinear Model Predictive Control with Cost Function Scheduling for a Wheeled Mobile Robot" (2022), where he introduces a novel MPC approach using scheduled quadratic stage costs to approximate sparse binary costs, improving control of nonlinear systems. Additionally, in "Multi-output Infinite Horizon Gaussian Processes" (2021), Lim advances online learning for robotics by extending GP-based methods to handle multi-output, streaming sensory data. His research is notable for its practical focus on bridging theoretical models with real-world robotic applications, making him a promising voice in the fields of learning-based control and autonomous systems.