zbMATH — the first resource for mathematics

Methodes ordinales et combinatoires en analyse des donnees. (Ordinal and combinatorial methods in data analysis). (French) Zbl 0641.68031
Summary: In this paper first we formulate some remarks on relations between discrete mathematics, computer sciences and data analysis. Then we present a set of methods using ordinal or combinatorial techniques. Boolean analysis and graph theory approach for binary data, tree representations (additive trees, n-trees), seriation methods for symmetric or not symmetric dissimilarity arrays, preferences aggregation procedures and consensus problems are studied. For each method we give a brief description and some bibliographic references concerning theoretical apsects and description of algorithms. Methods are classified according to the structures of the data processes: one or several arrays of type \({\mathbb{I}}\times {\mathbb{J}}\) or \({\mathbb{K}}\times {\mathbb{K}}\), with binary, ordinal or quantitative values.

68P05 Data structures
62-07 Data analysis (statistics) (MSC2010)
05C05 Trees
Full Text: Numdam EuDML