Scott M. Jordan
Postdoctor research at the University of Alberta studying reinforcement learning.
I am a Postdoctoral Fellow at the University of Alberta, co-advised by Professors Martha White and Adam White. I completed my Ph.D. in 2022 at the University of Massachusetts, where I was advised by Professor Philip Thomas. I primarily research techniques for solving sequential decision-making problems, focusing on using Reinforcement Learning techniques. My research interests center around answering the following high-level questions:
- What are the necessary properties for scaling reinforcement learning techniques to solve many subtasks?
- How can we make algorithms reliable and easy to use for non-experts, i.e., how can we make reinforcement learning algorithms work off the shelf?
- What principles should we, as researchers, follow when designing machine learning experiments? Similarly, how can we improve the quality of experiments in machine learning?
If you are interested in my research or collaborating, please reach out to me via email.
News
| May 15, 2024 | Paper accepted to the RLC showing the adaptive step sizes for policy gradient methods have to balance the exploration/exploitation trade-off. |
|---|---|
| Apr 16, 2024 | Position paper accepted to ICML arguing the benchmarking is limited and additional types of experimentation is needed. |
| Jan 1, 2024 | Agreed to be part of the RLC Program Chair Committee where we are piloting a new review process. |
| Dec 9, 2022 | Gave one of the opinion talks at the NeurIPS Deep RL Workshop. [Video] |
| Nov 1, 2022 | Started as a Postdoc at the University of Alberta working with Martha and Adam White. |