Supervised classification or pattern recognition is a method to solve decision problems in Social Sciences. It is organized on the basis of specific sets of predictor variables and the existence of classes known a priori. Based on a training sample, its main objective is to construct a classification rule in order to predict the class to which a new object belongs. Nowadays, the availability and efficacy of powerful computers have made possible many advances in this field, both in Statistics and Computer Sciences. In this section, different methods will be discussed and illustrated with the results obtained in several applications. The following topics will be dealt with: Parametric Discriminant Analysis, Non-parametric Discriminant Analysis, Logistic Discriminant Analysis, Neuronal Networks, Recursive Partitioning and Estimation of Error Rates.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education