The November 2015 UWIN seminar features Dr. Yoram Singer, a Senior Research Scientist at Google. He will be giving a talk entitled “Machine learning techniques for discovering spatio-temporal response of primate retinal ganglion cells”.
The seminar will be on Thursday, November 12, at 3:30pm in Health Sciences Building G-417. Refreshments will be served prior to the talk.
Dr. Singer runs a small research group at Google that focuses on machine learning principles. Prior to his position at Google he was an associate professor at the Hebrew University of Jerusalem, Israel. He was the co-chair of COLT in 2004 and NIPS in 2007. His work in machine learning has received a number of awards.
Seminar Abstract: We describe a low-rank model for predicting visual responses in ganglion cells. The model consists of temporal response filter which is joint for all cells from the same family and either a dedicated or tied spatial response filter. We cast the learning problem as a multilabel logistic regression. In order to find the spatial-temporal parameter we use a dual decomposition technique along with a new proximal stochastic gradient method. As a proof of concept we show how we reconstructed the typical temporal and spatial responses of ON and OFF Parasol cells. Joint work with Carlos Fernandez-Granda, Vineet Gupta, Jon Shlens, and E.J. Chichilnisky.