Research interestsMy research interests lie in the development of probabilistic models for the analysis of complex and structured data, the study of the theoretical properties of such models and the development of scalable inference algorithms. I am currently particularly interested in sparse networks, recommender systems and ranking systems.
- Statistical Machine Learning
- Bayesian Methods, Bayesian nonparametrics
- Monte Carlo methods
- Statistical Network modeling
- Fadhel Ayed, PhD student, Oxford
- Giuseppe Di Benedetto, PhD student, Oxford, co-supervised with Y.W. Teh
- Juho Lee, postdoc, Oxford
- Xenia Miscouridou, PhD student, Oxford, co-supervised with Y.W. Teh
- Jessica Sodjo, PhD student, INRIA Bordeaux, co-supervised
with A. Giremus, N. Dobigeon and J.F. Giovannelli
- Andrei Barbos, PhD student, University of Bordeaux (2014-2017), co-supervised with J.F. Giovannelli. Now image processing scientist at DxOMark.
- Adrien Todeschini, PhD student at INRIA Bordeaux (2013-2016), co-supervised with M. Chavent. Now data scientist at scorelab.io.
- Boris Hejblum, PhD student at University of Bordeaux (2011-2015), co-supervised with R. Thiébaut. Now assistant professor (maitre de conférence) at University of Bordeaux.
- Michele Pace, PhD student at INRIA Bordeaux (2009-2011), co-supervised with P. Del Moral. Now market risk expert at Generali Belgium.
SNetOC (Matlab package): Simulation and inference for Sparse and modular Networks with Overlapping Communities (SNetOC)
NPflow (R package): Bayesian Nonparametrics for Automatic Gating of Flow-Cytometry Data
BNPgraph (Matlab package): Simulation and posterior inference for sparse random graphs
BiiPS software: Bayesian Inference with Interacting Particle Systems
Polyatreetest (Matlab/Octave package): Bayesian nonparametric two-sample test with Polya tree priors
BNPPL (Matlab/Octave package): Bayesian inference in nonparametric (mixture of) Plackett-Luce models
HASI (Matlab package): Adaptive algorithm for low-rank matrix completion and collaborative filtering
BayesBT (Matlab/Octave package): Efficient Bayesian inference for generalized Bradley-Terry models
TeachingSB1b/SM2 Computational Statistics (Hilary Term 2018)
Department of Statistics
University of Oxford
24-29 St Giles
OX1 3LB, Oxford
Phone: +44 (0)1865 282 865