Cluster analysis applied to the Human Development Index (HDI) of Brazilian States

This study aims to compare the performance of each method (hierarchical and non-hierarchical) grouping formed by several HDI from 27 brazilian states, through cluster analysis technique. As well as determining how many states there are in group, thus specify which technique best represents data. Data Atlas Brasil 2013 were used relation 2010 HDI. For analysis, Mahalanobins matrix was with hierarchical method, data obtained, we applied simple linkage methods, complete, average, ward liaison a non-hierarchical K-means conphenetic correlation coefficient also measure degree fit between original similar matrices resulting simplification provided method. However, that complete link. When similarity HDI-R, HDI-L HDI-S variables considered this relationship groups connections different regions Brazil.