Análisis multivariante de datos

In this work a review of the main multivariate statistical methods is made, in order for researchers and professionals to acquire sufficient knowledge to properly use a set of statistical tools for multivariate analysis of data of interest for prediction and decision-making in the company, and as an essential statistical tool to find patterns of behavior in large big data databases. These tools are factor analysis, principal component analysis, cluster or conglomerate analysis, discriminant analysis, and logistic regression analysis. The book analyzes the relationships between the variables in a data set to summarize the information they contain, using a small set of theoretical variables. These theoretical variables, not observed, will be latent variables that extract the information from the observed variables to summarize and synthesize the information they contain. The objective of this reduction or synthesis is to facilitate the interpretation of the behavior of the population from which the data have been extracted. Similarities between individuals or cases in a data set are also studied to form classification groups with similar characteristics.