Dec, 2023
基于贝叶斯优化的课程学习提升自主驾驶深度强化学习方法的环境鲁棒性
Improving Environment Robustness of Deep Reinforcement Learning Approaches for Autonomous Racing Using Bayesian Optimization-based Curriculum Learning
Rohan Banerjee, Prishita Ray, Mark Campbell
TL;DR通过贝叶斯优化进行概率推理的课程学习方法在自主赛车避障领域表现优于基准的深度强化学习代理与手工设计的课程。