Personal tools
You are here: Home Events ANC/DTC Seminar: Guillaume Bouchard, Xerox Research Centre Europe (Host: Charles Sutton)

ANC/DTC Seminar: Guillaume Bouchard, Xerox Research Centre Europe (Host: Charles Sutton)

— filed under:

Convex methods for multi-view learning, link prediction and collective matrix factorization

What
  • ANC/DTC Seminar
When May 07, 2013
from 11:00 AM to 12:00 PM
Where IF 4.31/4.33
Add event to calendar vCal
iCal

Many applications involve multiple interlinked data sources, but existing approach to handle them are often based on latent factor models which are difficult to learn. At the same time, recent advances in convex analysis, mainly based on the nuclear norm (relaxation of the matrix rank) and sparse structured approximations, have shown great theoretical and practical performances to handle very large matrix factorization problems with non-Gaussian noise and missing data. In this talk, we will show how multiple matrices can be jointly factorized using a convex formulation of the problem, with a particular focus on:

· Multi-view learning: A popular approach is to assume that, both, the correlations between the views and the view-specific correlations have low-rank structure, leading to a model closely related to canonical correlation analysis called inter-battery factor analysis. We propose a convex relaxation of this model, based on a structured nuclear norm regularization.

· Collective matrix factorization: When multiple matrices are related, they share common latent factors, leading to a simple yet powerful way of handling complex data structures, such as relational databases. Again, a convex formulation of this approach is proposed.

Experiments on popular tasks such as data imputation, multi-label prediction, link prediction in graphs and item recommendation illustrate the benefit of the proposed approaches.