FJ-LMI Seminar

Seminar information archive ~07/24Next seminarFuture seminars 07/25~

Organizer(s) Toshiyuki Kobayashi, Michael Pevzner


16:00-   Room #117 (Graduate School of Math. Sci. Bldg.)
Maud DELATTRE (Université Paris-Saclay, INRAE)
Some contributions on variable selection in nonlinear mixed-effects models
[ Abstract ]
In the first part of this presentation, we will introduce the general formalism of nonlinear mixed effects models (NLMEM) that are specifically designed models to describe dynamic phenomena from repeated data on several subjects. In the second part, we will focus on specific variable selection technics for NLMEM through two contributions. In the first one, we will discuss the proper definition and use of the Bayesian information criterion (BIC) for variable selection in a low dimensional setting. High dimensional variable selection is the subject of the second contribution.

[1] Delattre, M., Lavielle, M. and Poursat, M.A. (2014) A note on BIC in mixed effects models, Electronic Journal of Statistics 8(1) p. 456-475.
[2] Delattre, M. and Poursat, M.A. (2020) An iterative algorithm for joint covariate and random effect selection in mixed effects models., The International Journal of Biostatistics 16(2), 20190082.
[3] Naveau, M., Kon Kam King, G., Rincent, R., Sansonnet, L. and Delattre, M. Bayesian high dimensional covariate selection in non-linear mixed-effects models using the SAEM algorithm. hal-03685060.
[ Reference URL ]