Reinforcement Learning Models
Mikko and Kyle discuss the conceptual difference between reinforcement learning and model-based approaches, highlighting the potential benefits of advancements in both methods. They explore the components of a model in reinforcement learning, including state space, action space, observation space, and probabilistic models. The conversation also touches on the contrast between deterministic games like chess and probabilistic models, emphasizing the need to work with expectations and distributions to achieve practical results.In this clip
From this podcast
Data Skeptic
Decentralized Information Gathering
Related Questions