Introduction to High-Dimensionality

Abstract

Abstract In an era in which the complexity and volume of data available in the field of machine learning is growing daily, feature selection plays an important role, helping to reduce the “high-dimensionality” of some problems. In this chapter, the problematics and characteristics of these “high-dimensional” datasets will be presented. Section 1.1 introduces the need for feature selection from the advent of Big Data. In Section 1.2, we outline the main applications that are promoting feature selection. Then, in Section 1.3, the inherent characteristics of some problems that may hinder the feature selection process are also discussed. Finally, we give an overview of the different chapters of this book in Section 1.4.