Event Details

In this talk I want to highlight what I personally think are some important open problems in Deep Learning. A big part of the talk will be focused on Continual Learning as one main open problem in our field. I will first try to motivate why Continual Learning is an important open problem. Secondly I will try to highlight the fluidity of the term, and be slightly more precise with what aspects of the problem I consider to be crucial. In the second half of my talk I will go over a few interesting recent observations, highlighting the role implicit biases can have in learning,  and therefore the importance of understanding these biases. Overall my goal is to provide a cohesive story about some interesting open problems that we need to solve in order to make further progress, as well as introducing a few of my recent works in order for the audience to get some sense of the topics I’m interested in, that we could pick up in the QA section of the talk, or afterwards in 1:1 discussions.

Speaker Bio

Razvan did his PhD in Montréal with Yoshua Bengio. He has been a prolific researcher that made a variety of impactful contributions to Deep Learning including optimization, recurrent models, graph neural networks, meta-learning, and continual learning. He is also actively promoting AI research and education as a main organizer of Conference on Life-long Learning Agents (CoLLAs) , Eastern European Machine Learning Summer School (EEML) and different workshops at NeurIPS, ICML and ICLR.

YOU ONLY NEED TO REGISTER ONCE TO ATTEND THE ENTIRE SERIES – We will send you email announcements with details of the upcoming speakers.

Register in advance for this webinar: https://usc.zoom.us/webinar/register/WN__0VhakI6Q6i3JsasdmNWcA

After registering, you will receive an email confirmation containing information about joining the Zoom webinar.

If speaker approves to be recorded for this AI Seminar talk, it will be posted on our USC/ISI YouTube page within 1-2 business days: https://www.youtube.com/user/USCISI.

Host: Mohammad Rostami, POC: Pete Zamar