Speaker: Prof. Florian Yger (Université Paris-Dauphine, France)
Title : Topics in recommendation, matrix factorization and completion
Abstract : After introducing the problem of recommending systems, this presentation covers the basics of collaborative filtering and content-based filtering. Then, it presents a novel approach (bridging the gap between those two previous approaches) which enables cold-start in a principled way for collaborative filtering system. Finally, the related problem of covariance matrix completion and its use in the context of BCI data will be introduced.
|Date||July 14, 2017 (Fri) 10:30 - 12:00|