Extra Colloquium Mathematics, Anani Lotsi

Join us for coffee and tea at 16.00 p.m.

Date:                       Thursday, February 13^th 2014

Speaker:                  Anani Lotsi

Room:                      5161.0165 (Bernoulliborg)

Time:                        16.15

Title: State space and graphical models for estimating networks dynamics

Abstract:

Genomic systems have become complex. Networks reconstruction are seen as an attractive paradigm of genomic science. My thesis is concerned with estimating networks dynamics from inhomogeneous samples and time complex dynamics. We developed two new methods to deal with this scenarios.

In this talk I will present  two models we have used for networks estimation: graphical mixture models and state space models  and  suggested two novel methods of inferences namely penalized Gaussian graphical mixture models  and penalized state space models. Given the incompleteness nature of information, we propose the Expectation-Maximization  algorithm  as solution to such incomplete data problems.

I will illustrate general consistency results for the penalized maximum likelihood estimator in the Gaussian graphical mixture model and show the performance  of our results in  parameter consistency as well as in graph selection consistency. For the complex time dynamics data, we have built an input dependent state space models  and employed a penalized maximum likelihood inference as estimation method.

Colloquium coordinators are Prof.dr. A.C.D. van Enter (e-mail : A.C.D.van.Enter@rug.nl) and
Dr. A.V. Kiselev (e-mail: a.v.kiselev rug.nl )

http://www.rug.nl/research/jbi/news/colloquia/mathematics-colloquia/