CMStatistics 2017: Start Registration
View Submission - CMStatistics
Title: Variable selection for model-based clustering: Application in human population genomics Authors:  Matthieu Marbac - ENSAI (France) [presenting]
Mohammed Sedki - Paris-Sud University, Inserm, Pasteur, UVSQ (France)
Abstract: Two approaches are considered for selecting variables in latent class analysis. The first approach consists in optimizing the BIC with a modified version of the EM algorithm. This approach simultaneously performs both model selection and parameter inference. The second approach consists in maximizing the MICL, which considers the clustering task, with an algorithm of alternate optimization. This approach performs model selection without requiring the maximum likelihood estimates for model comparison, then parameter inference is done for the unique selected model. Thus, both approaches avoid the computation of the maximum likelihood estimates for each model comparison. Moreover, they also avoid the use of the standard algorithms for variable selection which are often suboptimal (e.g. stepwise method) and computationally expensive. The case of data with missing values is also discussed. The interest of both proposed criteria is shown on an application in human population genomics problem. Data set describes 1300 patients by 160000 variables.