Combinatorial Inference in Geometric Data Analysis

Brigitte Le Roux develops a set of statistical inference techniques for geometric data analysis, built on a combinatorial foundation. Standard statistical models rest on assumptions that can rule out certain questions, and these methods are designed to address exactly those questions instead.

The book walks through the full methodology and illustrates its uses with detailed case studies drawn from applied work. Readers who want to try the techniques themselves will find R code included for implementing them.

The text is aimed at researchers and students of multivariate statistics, along with applied researchers from a range of scientific fields. It could also serve as the basis for a specialized course at the master's or doctoral level.