The MasterClass with Matthew Taylor will involve 2 one hour lectures with plenty of time for Q&A!
Lecture 1: Reinforcement Learning in the Real World
March 22nd, 12:30pm ET
Reinforcement learning (RL) has had many successes, from AlphaGo to StarCraft 2 and data center cooling. Unfortunately, many real-world business tasks are not easily defined and can be difficult to formulate so that RL can successfully solve them. Furthermore significant amounts of time and/or data may be required to reach acceptable performance, potentially making learning infeasibly slow or expensive. This talk will:
- Briefly remind the audience what RL is;
- Discuss how to effectively identify and formulate an RL problem, including concrete use cases and;
- Identify ways to leverage existing knowledge from other agents, programs, or people to help RL agents to improve their performance.
Lecture 2: Centaurs and Augmented Intelligence: Human-AI Systems in 2021
March 29th, 12:30pm ET
Would you want a computer to decide whether you needed an ambulance when you called 911? For the foreseeable future, machine learning and artificial intelligence systems will need to incorporate humans in high-stakes environments. Furthermore, even if your end goal is a completely autonomous solution, it may be much easier to have an initial deployment where the AI is assisting a human. This talk will discuss
- When and why human-AI systems can outperform human-only or AI-only approaches by leveraging the unique abilities of both humans and AIs,
- State of the art methods for successful combinations, and
- Showcase successful examples to help you identify potential centaur approaches in your business.
Matthew E. Taylor received his doctorate from the University of Texas at Austin in the summer of 2008, supervised by Peter Stone. He then completed a two-year postdoctoral research position at the University of Southern California with Milind Tambe and spent 2.5 years as an assistant professor at Lafayette College. He was then an assistant professor at Washington State University where he held the Allred Distinguished Professorship in Artificial Intelligence. In 2017, he temporarily left academia to help start an artificial intelligence lab in Edmonton, Alberta, with Borealis AI, the artificial intelligence research lab for the Royal Bank of Canada, where, among other things, he contributed to a deployed reinforcement learning agent for stock trading.
He is now a tenured associate professor in computer science at the University of Alberta, a Fellow-in-Residence at the Alberta Machine Intelligence Institute, and remains an adjunct professor at Washington State University. He has been a PI or co-PI on over $7M USD in competitively awarded research funding from federal, state, and industrial sources, including a National Science Foundation CAREER award and a Canada CIFAR AI Chair. He has (co-)supervised 7 graduated PhD students and 5 MS students in the Intelligent Robot Learning Lab, as well as published over 120 peer-reviewed papers in conferences and journals. His current fundamental and applied research interests are in reinforcement learning, human-in-the-loop AI, multi-agent systems, and robotics.