Métodos y procesos sofisticados de reducción de dimensión y selección de variables para un modelo de respuesta binaria.

Abstract

The current challenge for researchers and professionals is the rate of development regarding data collected. The manipulation of this large amount of data is currently considered an open and booming topic. Part of the information that is collected is usually irrelevant and unnecessary, which motivates data analysts to use technological resources that exclude this information, achieving results referring to some phenomenon. In this paper, methods and techniques for variable selection and dimension reduction are proposed to overcome the problem of handling large amounts of data. On the one hand, it goes deeper into machine learning algorithms and then into unusual selection techniques. These methods are applied and finally a validation of the results is carried out, concluding an adequate proposed model. The database used in this project comes from a private consultancy.