Vector operation-based decision class joint set up and down approximation acquisition method

Abstract

The invention discloses a vector operation-based decision class joint set up and down approximation acquisition method. The method comprises the following steps: firstly, representing a dominant set and a disadvantage set of each discourse domain object as a dominant row vector and a disadvantage row vector respectively; then, respectively calculating products of corresponding elements of the dominant row vector, the inferior row vector and the decision row vector of each discourse domain object to obtain two row vectors; constructing two generalization decision row vectors, namely a minimum generalization decision row vector and a maximum generalization decision row vector, according to the minimum value and the maximum value of the elements in the two row vectors respectively, wherein the decision values are respectively composed of minimum decision values of all objects in a dominant set of each object in the discourse domain and maximum decision values of all objects in a disadvantage set of each object; and finally, obtaining upper and lower approximations of the decision class joint set according to the four different forms of section vectors of the generalization decision vectors. According to the method, the problem of simple and intuitive calculation of upper and lower approximation of the decision class joint set in the dominant decision system is solved from the perspective of vectors.

Description

technical field [0001] The invention belongs to the field of dynamic knowledge discovery based on the dominant relational rough set method, and specifically relates to a method for obtaining upper and lower approximations of decision-making union sets based on vector operations. Background technique [0002] The classic rough set method proposed by Pawlak is an efficient mathematical tool to deal with problems with inconsistent, imprecise and fuzzy information, and has been successfully applied in many fields such as pattern recognition, data mining and knowledge discovery. [0003] The calculation of conceptual upper and lower approximation sets is a hot issue in rough set research, and it is the basis of rule extraction and attribute reduction (feature selection) in subsequent research. Because the classic rough set method does not consider the preference relationship between attribute values, it cannot deal with the preference information (ordinal information) in the attr...

AI Technical Summary

Problems solved by technology

However, according to the original definition of the upper and lower approximation of the decision-making joint set in the dominant relational rough set method, the calculation efficiency of directly obtaining the approximate set is low and the calculation process is complicated.

Images