[Eecs_phd] ACM Presentation on Reinforcement Learning-Today 11/9 ARC 116 7:30pm

[Hunter, Tiffany]

Hi Everyone!

This week grad student Patrick Gray will be presenting on an area within machine learning called "reinforcement learning." This presentation will include the fundamental ideas and math behind the topic and its applications to the real world. If you're interested in expanding your knowledge of machine learning, join us this Wednesday (11/9) at 7:30pm in ARC 106!



Abstract:

"Reinforcement learning is a type of machine learning in which computers automatically learn how to act within a particular environment so as to maximize a numerical reward signal. The computers are never told which actions to take, but instead must discover for themselves, through trial and error, the most rewarding actions. For example, a self-driving car implemented with a reinforcement learning algorithm will automatically learn how to properly adjust its steering wheel through a reward system that reinforces stability and deters swerving. Although this idea seems rather simple, it works quite well in practice. All in all, reinforcement learning is an exciting area of artificial intelligence and will undoubtedly power many of the automated technologies of the near future.



On Wednesday, I will walk you through the basics of reinforcement learning, from the common nomenclature to the fundamental (and not at all daunting) math. To help solidify your understanding of the basics, I will also present to you a comprehensible solution to a simple, yet important, real-world reinforcement learning problem. The presentation will conclude with a survey of some very impressive reinforcement learning systems that trained themselves to fly helicopters, diagnose diseases, and even play some classic video games."

- Patrick Gray



We hope to see you all there!



Sincerely,

Catherine Baugher

OU ACM Secretary
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://listserv.ohio.edu/pipermail/eecs_phd/attachments/20161109/9dcac13a/attachment.html