Aplicación de técnicas estadísticas multivariadas en perfilación y segmentación

Abstract

Objective.To use the information about the factors identified, by means of correspondence analyses to determine the behavior of the variables to be used to then make a natural segmentation, thus facilitating the interpretation and presentation of results. Materials and methods. We present a new profiling and segmentation approach by using correspondence analyses and classification. Results. By using this methodology we can optimize the determination of segments and the profiling of a set of variables. This procedure allows researchers of different disciplines to present and interpret the results of their research efficiently. Conclusions.  The procedure here described enables a simple and quick interpretation of the analysis on many variables, rendering its use and implementation easy. Key words: profiling, segmentation, multivariate techniques, correspondences.