Deep Reinforcement Learning
Deep reinforcement learning (DRL) aims to train agents to make optimal decisions in complex environments by learning through trial and error. Current research focuses on improving DRL's robustness, sample efficiency, and interpretability, often employing architectures like Proximal Policy Optimization (PPO), deep Q-networks (DQNs), and graph neural networks (GNNs) to address challenges in diverse applications such as robotics, game playing, and resource management. The resulting advancements have significant implications for various fields, enabling the development of more adaptable and efficient autonomous systems across numerous domains.
Papers
Towards Multi-agent Policy-based Directed Hypergraph Learning for Traffic Signal Control
Kang Wang, Zhishu Shen, Zhenwei Wang, Tiehua Zhang
Soft Actor-Critic with Beta Policy via Implicit Reparameterization Gradients
Luca Della Libera
Enhancing Socially-Aware Robot Navigation through Bidirectional Natural Language Conversation
Congcong Wen, Yifan Liu, Geeta Chandra Raju Bethala, Zheng Peng, Hui Lin, Yu-Shen Liu, Yi Fang
SpecGuard: Specification Aware Recovery for Robotic Autonomous Vehicles from Physical Attacks
Pritam Dash, Ethan Chan, Karthik Pattabiraman
Earth Observation Satellite Scheduling with Graph Neural Networks
Antoine Jacquet, Guillaume Infantes, Nicolas Meuleau, Emmanuel Benazera, Stéphanie Roussel, Vincent Baudoui, Jonathan Guerra
Localized Observation Abstraction Using Piecewise Linear Spatial Decay for Reinforcement Learning in Combat Simulations
Scotty Black, Christian Darken
Robust Iterative Value Conversion: Deep Reinforcement Learning for Neurochip-driven Edge Robots
Yuki Kadokawa, Tomohito Kodera, Yoshihisa Tsurumine, Shinya Nishimura, Takamitsu Matsubara