Safe Reinforcement Learning for Autonomous Driving via Expert Advice
Proposing an uncertainty-aware framework to guide exploration in reinforcement learning for autonomous vehicles, helping to avoid collisions during training.
An uncertainty-aware framework is proposed to guide exploration in reinforcement learning (RL) for autonomous vehicles. Expert advice is only triggered when uncertainty metrics exceed a threshold, helping to avoid collisions during training in CARLA.
Why It Matters
Minimizes risks when deploying machine learning in safety-critical physical systems like autonomous vehicles.
