REVISTA DYNA NEW TECHNOLOGIES

Key words:

Conjuntos aproximados, Algoritmos de agrupamiento conceptual, Testores típicos, Clasificación no supervisada, Rough Sets, Conceptual Clustering Algorithms, typical Testores, not supervised Classification

Article type:

ARTICULO DE INVESTIGACION / RESEARCH ARTICLE

Section:

RESEARCH ARTICLES

The rough sets theory and the theory of conceptual clustering do not contend but rather complement each other; actually, they are both independent approaches used for solving problems such as data reduction, dependency discovery among data, estimate of data significance, generation of decision or control algorithms from the data obtained, approximate data classification, discovery of similarities or differences within the data and pattern discovery, among others. Specifically, both theories have had an interesting application in the fields of medicine, business, engineering design, and meteorology, in vibration analysis, conflict analysis, image processing, voice recognition, characters recognition, decision analysis and so forth.
Concerning the theory of classical groups, it is well-known that these are not defined, but are actually specified, and that there are two ways to do it: an extensional way, which means reveal its constituents, and an intentional way, which means revealing the properties that characterizes the group, that satisfy the elements in the group.

This paper suggests a method for determining the lower and upper approximations sets, for the approach based on the similarity relationships of the rough sets theory using the conceptual clustering with a combinatorial logic vision.