Based upon guidance provided by the University System of Georgia, all Georgia Tech sponsored events through July 31, including athletics competitions, are cancelled, postponed or will move to a virtual format.

Tuesday, March 24 2020
11:00am - 12:00pm
For more information:

Anna Stroup

Add To My Calendar
CSE Faculty Candidate Seminar - Zhiting Hu

Seminar Date: Tuesday, March 24, 2020

Seminar Time: 11:00 am

How to Attend:

To join the meeting on a computer or mobile phone:

Meeting ID: 716 632 207

  • Room System
  • or

Talk Title:  Towards Training AI Agents with All Types of Experiences via a Single Algorithm

Talk Abstract: Training AI agents for complex problems, such as controllable content generation, requires integrating all sources of experiences (e.g. data, constraints, cost, information from other tasks) in learning. Past decades of research has led to a multitude of learning algorithms for ingesting different experiences. However, creating solutions based on such a bewildering marketplace of algorithms demands strong ML expertise and bespoke innovations. This talk will present an alternative approach to creating solutions from a unifying perspective. I will show that many of the popular algorithms in supervised learning, constraint-driven learning, reinforcement learning, etc, indeed share a common succinct formulation and can be reduced to a single algorithm that enables learning from different experiences in the same way. This allows us to create solutions by simply plugging arbitrary experiences in learning, and to enable new learning capabilities by repurposing off-the-shelf algorithms

Bio: Zhiting Hu is a Ph.D. student in the Machine Learning Department at CMU. He received his B.S. from Peking University. His research interests lie in the broad area of machine learning. His research was recognized with best demo nomination at ACL2019, best paper award at ICLR 2019 DRL workshop, outstanding paper award at ACL2016, and IBM Fellowship.

Host: Srijan Kumar