JC Schoeman is a Senior Lecturer in the Department of Electrical and Electronic Engineering at Stellenbosch University. He was first appointed to the department in 2019 and was promoted to Senior Lecturer. He obtained his PhD from Stellenbosch University in 2021. The dissertation, completed at the Electronic Systems Laboratory under the supervision of Corné van Daalen and Johan du Preez, was titled Degenerate Gaussian factors for probabilistic inference.
His research focuses on the intersection of robotics and machine learning, with particular emphasis on decision theory, reinforcement learning, inference and optimisation, autonomous vehicles, and games. Schoeman’s work aims to develop autonomous systems with optimal decision-making capabilities by unifying concepts from computer science, control systems, and signal processing. Key publications include Degenerate Gaussian factors for probabilistic inference in the International Journal of Approximate Reasoning in 2022, Partial End-to-end Reinforcement Learning for Robustness Against Modelling Error in Autonomous Racing as an arXiv preprint in 2023, LABCAT: Locally adaptive Bayesian optimization using principal-component-aligned trust regions in Swarm and Evolutionary Computation in 2025, Augmenting the action space with conventions to improve multi-agent cooperation in Hanabi in Autonomous Agents and Multi-Agent Systems in 2025, Lossy compression of observations for Gaussian process regression in MATEC Web of Conferences in 2022, and Credit-cognisant reinforcement learning for multi-agent cooperation as an arXiv preprint in 2022. His contributions centre on probabilistic inference, Bayesian optimisation, and multi-agent reinforcement learning applications.