Many problems in different scientific domains can be described through statistical models that relate the sequential observed data to a hidden process through some unobserved parameters. In the Bayesian framework, the probabilistic estimation of the unknowns is represented by the posterior distribution of these parameters. However in most of the realistic models, the posterior is intractable and must be approximated. Importance Sampling (IS)-based algorithms are Monte Carlo methods that have shown a satisfactory performance in many problems of Bayesian inference.
In this project, we will apply and extend recent adaptive IS-based methods for probabilistic inference in complex non-linear high-dimensional systems. Novel adaptation schemes will be explored in order to overcome current limitations of such techniques. Many applications can be benefited from the development of these methodologies. In particular, we will apply the Bayesian inference techniques in problems such us tracking of space debris or prediction of the evolution of a biological system related to the formation and growth of cancer stem cells.
The thesis will take place as soon as possible for a duration of 5/6 months, at Télécom Lille in Villeneuve d’ascq (http://www.telecom-lille.fr) and CRIStAL-SIGMA Team (http://www.cristal.univ-lille.fr/?rubrique29&eid=30). The student will be supervised by:
Please send us a CV + motivation letter to apply for this position.
Perspective: a PhD funding on this topic might be possible, starting from September/October 2017
(c) GdR 720 ISIS - CNRS - 2011-2015.