The National Science Foundation awarded a $600,000 Project Grant to Princeton University under the Computer and Information Science and Engineering program (CFDA 47.070).
The grant will support research toward developing a mathematical foundation for deep reinforcement learning over the period from October 1, 2022 to September 30, 2026. Specifically, the project aims to bridge current gaps in theoretical reinforcement learning and deep neural networks by investigating guarantees achievable by reinforcement learning policies under different problem structures. Researchers will also examine how controlling neural network complexity can yield sample-efficient algorithms and design computationally efficient algorithms by leveraging tools from optimization and control theories. The project integrates education and outreach activities including student mentorship and developing new course materials.
Generated 1/7/24, 11:28 AM