BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Optimal Link Prediction with Matrix Logistic Regression - Quentin 
 Berthet (University of Cambridge)
DTSTART:20180116T140000Z
DTEND:20180116T144500Z
UID:TALK97732@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:We consider the problem of link prediction\, based on partial 
 observation of a large network and on covariates associated to its vertice
 s. The generative model is formulated as matrix logistic regression. The p
 erformance of the model is analysed in a high-dimensional regime under str
 uctural assumption. The minimax rate for the Frobenius norm risk is establ
 ished and a combinatorial estimator based on the penalised maximum likelih
 ood approach is shown to achieve it. Furthermore\, it is shown that this r
 ate cannot be attained by any algorithm computable in polynomial time\, un
 der a computational complexity assumption. (Joint work with Nicolai Baldin
 )
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR