Article

Enhancing Robotic Motion Planning in Deep Reinforcement Learning via Task Decomposition

This study proposes a task decomposition framework for deep reinforcement learning (DRL) that combines transfer learning (TL) with reward redesign (RR) to improve learning efficiency and robustness in obstacle environments.

Part of the event: New Framework Combines Transfer Learning and Reward Redesign to Improve DRL Motion Planning

This page provides a summary and source link only.