Compactness Classification for Geographic Zones

With several goals, one of the classical problems in population studies is the classification of zones or variables for a metropolitan area. The most widely known non-supervised classification algorithms present drawbacks in zonification problems since the grouping process does not allow manual control of the variables. In the population analysis problems involved with zonification it is common to require the specification of certain bounds for some indicators (variables within a given interval) to create groups and thus determine the level of group membership. As a result the implementation of a process to classify and extract population variables from Mexico´s XII national population census, this work describes a compact and homogeneous classification algorithm it has been implanted