Uma análise de regressão logística usando componentes principais

Abstract

In this Bachelors dissertation we studied, in the context of logistic regression models, the case in which the number of covariables is large and the covariables are correlated. In this situation, we use the principal component analysis technique in order to reduce the number of the variables (dimension) involved in the data set. We present two examples of logistic regression and one of principal component analysis, as well as a comparison between an application with both methodologies, applied jointly, and an application using only logistic regression.