Second semester
Multivariate data analysis
Objectives
Detection of homogeneous subpopulations ; Mastery of method implementation algorithms ; Critical discussion of results and method selection ;
Course outline
Introduction to unsupervised classification. Difficulties, interpretations, and issues ; Geometric classification methods: k-means, hierarchical ascending clustering ; Mixture models. EM algorithm ; Probabilistic classification methods: mixture model classification ; Analysis of results, methodological choices, and application contexts ;
Prerequisites
Linear algebra, Descriptive statistics, Probability