Reinforcement Learning introduces fundamentally different both algorithmic and systems challenges than the more well-known supervised learning paradigm. Reinforcement Learning is computationally harder because we are dealing with an agent learning in a dynamic environment while also computing simulations from that environment. In this tutorial, we will go from the fundamentals of applied Reinforcement Learning to applications in real-world settings like finance, supply chain optimization, and robotics. We will also build distributed reinforcement learning algorithms using newly introduced framework Ray, coming out of Rise Lab at Berekely. After the tutorial, attendees will have a strong foundational understanding of RL foundations; areas where it can be applied to it; and strong facility with Ray to prototype RL algorithms on their own.

Time and location will be posted when available.

Return to: Hands-on Tutorials