Partitioning the object-attribute space for data mining based on the merger of object elements

Abstract

The research of partitioning the object-attribute space belongs to the domain of interpretative structural modeling and it is one of the basic problems in the data mining field. Firstly this paper proposes and demonstrates the Subsystem Judgment theorem. In order to solve the problem above, an algorithm that reduces the scale and the dimension of the original data through partitioning the object-attribute space based on the merger of object elements is put forth. In the last part of the paper, a numerical value example is provided to show the whole process of the method.