FWO project - Iterative and multi-level methods for Bayesian multirelational factorization with features
From 01-01-2016 to 31-12-2019
Description
Many machine learning problems (classification, clustering, etc.) can be formulated as afactorization of an incompletely filled matrix where the goal is to predict the unknown values. These methods have been successful in large-scale recommender systems, like the Netflix challenge aimed at predicting movie preferences of users from 200 million movie ratings from half a million users. In particular, we are now generalizing the Bayesian Probabilistic Matrix Factorization by developing a method - Bayesian multirelational factorization with features - that can simultaneously factorize multiple relations (i.e., multiple matrices) and also incorporate additional entity-level features.While basic approaches to matrix factorization rely on classical optimization, our approach relies on Bayesian Markov Chain Monte Carlo, specifically Gibbs sampling. In fact, our approach can be reformulated as a sequence of linear system solves with many right-hand sides. If the dimensionality of the features is high, then solving the corresponding linear systems becomes the main computational bottleneck. This project aims at scaling up the method to very large sparse problems by reducing of the costs associated with the linear system solver. The bottlenecks we will tackle are (1) the speed and number of iterations of the Gibbs sampler during burn-in and (2) the number of iterations of the linear solver (which increases further with the size of the data set)
Team
- Karl Meerbergen, Co-promoter (External)
- Dirk Roose, Co-promoter (External)
- Yves Moreau, Promoter
Financing
Funding: FWO - Research Foundation - Flanders
Program/Grant Type: FWO Research Grant - FWO Research Grant
Events
2/09/2024:
PhD defense - Martijn Oldenhof
Machine Learning for Advanced Chemical Analysis and Structure Recognition in Drug Discovery
3/09/2024:
Meet the Jury Igor Tetko on Advanced Machine Learning in Drug Discovery
12/09/2024:
Multimodal analysis of cell-free DNA for sensitive cancer detection in low-coverage and low-sample settings
Seminar by Antoine Passemiers
News
STADIUS Alumni Herman Verrelst – new CEO of Biocartis
08 June 2017
Herman Verrelst, the founder of KU Leuven spin-off Cartagenia, who has been working in Silicon Valley, US for the last few years will be returning to Belgium to follow the steps of Rudi Pauwels as CEO of the Belgian diagnostic company, Biocartis.
Supporting healthcare policymaking via machine learning – batteries included!
29 May 2017
STADIUS takes the lead in the data analytics efforts in an ambitious European Project MIDAS.
Marc Claesen gives an interview about his PhD for the magazine of the Faculty of Engineering Sciences "Geniaal"
10 February 2017
Did you know that in Belgium approximately one third of type 2 diabetes patients are unaware of their condition?
Joos Vandewalle is nieuwe voorzitter KVAB
09 October 2016
Op 5 oktober 2016 heeft de Algemene Vergadering van de Academie KVAB Joos Vandewalle verkozen tot voorzitter van de KVAB.