Editorial for issue 2/2019


This issue 2 of volume 13 (2019) of the journal Advances in Data Analysis and Classification (ADAC) contains 10 articles that deal with discrimination (quadratic classifiers, trees) and clustering (e.g., mixtures, modal clustering), variable selection, agreement measures for rankings, missing data in PCA, symmetry in contingency tables, astronomy data and household clustering.

The first paper, written by José R. Berrendero and Javier Cárcamo, is entitled “Linear components of quadratic classifiers” and shows how a given quadratic classification rule (e.g., Mahalanobis classifier for normal distributions, a Bayesian rule, support vector machine) can be approximated by a combination of linear classifiers that are more easily to execute and interpret. The paper considers the two-class situation, approximates—in its simplest case—the underlying quadratic decision function by the product of a pair of linear functions (up to terms obtained by a suitable spectral decomposition) in the...


Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Personalised recommendations