Assistant professor Amy Zhang of Texas ECE received an Army Research Office award for her research project “Combinatorial Generalization through Search-based Planning and Reinforcement Learning” to advance the ability of cyber physical systems to navigate unknown environments, manipulate objects and team with other agents in the open world.

This requires reuse of knowledge and skills to generalize new environments and tasks at inference time, where current methods are still lacking. Her research group will investigate how to combine search-based methods into deep RL algorithms to increase the problem-solving effectiveness of real-world agents. 

Zhang’s research focuses on theory and algorithms for sequential decision-making problems, with an emphasis on reinforcement learning, self-supervised learning, and representation learning, with a concentration on improving robustness, generalization, and sample efficiency.