Danning Yu

About

Danning Yu is a researcher whose work lies at the intersection of deep reinforcement learning and partially observable systems. Her key contributions address a critical gap in the field: the assumption that state information is fully observable, which limits the applicability of many deep reinforcement learning algorithms in real-world scenarios. In her most cited work, "Deep Q-Network with Predictive State Models in Partially Observable Domains" (2020), Yu introduced a novel framework that integrates predictive state representations with deep Q-networks, enabling agents to effectively handle continuous, noisy observations. This work, while early in its citation trajectory, has laid important groundwork for more robust decision-making in complex, uncertain environments. Yu's research is particularly relevant for applications in robotics, autonomous navigation, and sensor-based control systems. By addressing the challenges of partial observability, she is helping to bridge the gap between theoretical reinforcement learning and practical deployment. Her contributions are gaining recognition as the field increasingly turns toward real-world, partially observable problems.

Research Focus

Key Achievements

Top Papers

1. 1
   Deep Q-Network with Predictive State Models in Partially Observable Domains

   2 citations · 2020

Key Collaborators

Contact & Links