Functional Model-Based Clustering

• Activity Number: 284
• Type: Topic Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistical Computing
• Author(s): Alejandro Murua*+ and Folly Adjogou and Wolfgang Raffelsberger
• Companies: University of Montreal and Universite de Montreal and Universite de Strasbourg
• Keywords: Functional analysis ; classification ; gene expression ; time-course data

Abstract

We develop a flexible model for the analysis and clustering of complete or sparse time-course or longitudinal data. The model combines functional analysis and model-based clustering. The functional modeling is based on splines. The main data groups are modeled as arising from clusters in the space of spline coefficients (the factors). The clusters are modeled by a mixture of Student's t-distributions whose degrees of freedom are unknown. The model is embedded into a Bayesian framework. We develop an approximation of the marginal log-likelihood MLL that allows us to do perform an MLL based model selection. Our criterion compares favorable with other popular criteria such as AIC and BIC. We also consider an extension of our model to curves in multiple dimensions. We will show some applications to gene expression data.