On aggregation of groups and categories in contingency tables

E. Carrizosa, V. Guerrero, D. Romero Morales

Extracting knowledge from data, such as dependencies, global underlying patterns or unusual behaviors, has become a crucial task for analysts to improve decision making. Nevertheless, the increase in data complexity has made, in some cases, the classic statistical models obsolete, and more sophisticated frameworks are thus needed. In this context, Mathematical Optimization plays an important role, both developing new models and algorithmic approaches as well as creating new frameworks, which gain insight into specific datasets’ features and cope with nowadays requirements. In this talk, we discuss how mixed integer nonlinear programming can be used to draw conclusions about statistical dependencies between categorical variables when different categories are allowed to be grouped.

Keywords: Mathematical Optimization; Contingency tables; Statistical dependence

GT4-1 Multivariate Analysis and Classification
September 3, 2019  3:30 PM
I3L10. Georgina Blanes building

J. A. Martín-Fernández, V. Pawlowsky-Glahn, J. J. Egozcue

E. del Barrio, H. Inouzhe Valdes, J. Loubes, C. Matrán, A. Mayo-Íscar