
Barrier Functions Inspired Reward Shaping enhances training efficiency and safety in RL.


coremsg

barrier-functions-inspired-reward-shaping-for-reinforcement-learning-a-novel-safety-oriented-framework


Barrier Functions Inspired Reward Shaping for Reinforcement Learning: A Novel Safety-Oriented Framework


title_rewrite


This paper introduces SASR, a novel self-adaptive reward shaping method for reinforcement learning that leverages success rates derived from historical experience to enhance learning in environments with sparse rewards.


self-adaptive-success-rate-based-reward-shaping-for-reinforcement-learning-in-sparse-reward-environments


Self-Adaptive Success Rate Based Reward Shaping for Reinforcement Learning in Sparse Reward Environments



ORSO is a novel approach that accelerates reward design in reinforcement learning by framing it as an online model selection problem, efficiently identifying effective shaping reward functions without human intervention.


orso-using-online-reward-selection-and-policy-optimization-to-accelerate-reward-design-in-reinforcement-learning


ORSO: Using Online Reward Selection and Policy Optimization to Accelerate Reward Design in Reinforcement Learning
