The PRINCOMP.lng Model

Principal Components Analysis

View the model
Download the model

The problem:

1) Given a data set in which variables display a significant amount of multicollinearity, we want to form new variables that are uncorrelated among themselves (i.e. orthogonal),

2) Given a large data set, we want to reduce the dimension of data to unity (e.g. composite index construction) with the minimum loss of information;

Based on a version by Eren Ocakverdi;


Statistics | Principal Components Analysis | Factor Analysis | PCA |