Probabilistic matrix factorization for reconstruction of missing data
Add to your list(s)
Download to your calendar using vCal
If you have a question about this talk, please contact Zoubin Ghahramani.
I will present our recent research on using probabilistic models for reconstruction of missing data. The first application considered is the task of collaborative filtering which is prediction of users’ preferences by learning past user-item relationships. We have developed a computationally efficient implementation of the variational Bayesian principal component analysis and used it in the Netflix prize problem. The second application is historical reconstruction of climate fields from available observations. I will present a matrix factorization model, which we call variational Gaussian-process factor analysis, that is suited for modeling spatio-temporal data.
This talk is part of the Machine Learning @ CUED series.
This talk is included in these lists:
Note that ex-directory lists are not shown.
|