Subset grouping semi-supervised fuzzy clustering method

Abstract

The invention relates to a subset grouping semi-supervised fuzzy clustering method, which comprises the following steps that: collecting the grouping results of a user for different subsets, obtaining a certain quantity of subset grouping information as clustering guidance to obtain the nearest neighbor set and the further neighbor set of objects in the subsets, and adding penalty terms into a fuzzy clustering target function to realize a purpose that the category of each object in the subset is similar to the category of the object in the neighbor set and is different from the category of the object in the further neighbor set to finish clustering. On the basis of a pairwise relationship between the objects, the fuzzy clustering not only considers decomposition errors but also considers a situation of no supervision information and a situation that subset grouping monitoring information is added, the decomposition errors are considered while matrix decomposition, meanwhile, a difference between the grouping of the objects in the subset and the grouping of the objects in the neighbor set is considered, a difference with the grouping of the objects in the neighbor set is increased, a satisfactory effect can be achieved without an overhigh number, clustering is quick in practical application, efficiency is high, and manual cost is low.

Description

technical field [0001] The invention belongs to the technical field of electrical digital data processing, in particular to a subset grouping semi-supervised fuzzy clustering method based on machine learning and data mining and introducing auxiliary information. Background technique [0002] In data processing and analysis problems in many fields, it is necessary to use clustering algorithms to group samples in a data set so that similar samples are classified into the same group, and then quickly browse the internal structure of the entire data set based on the grouping results , analysis and processing. [0003] The advantage of fuzzy clustering over hard clustering is that it introduces the concept of fuzzy membership with the help of fuzzy set theory, so that it can describe the overlap between classes naturally. In order to improve the accuracy of traditional unsupervised clustering methods, semi-supervised clustering using a small amount of supervised information is p...

AI Technical Summary

Problems solved by technology

[0005] The technical problem solved by the present invention is that in the prior art, the semi-supervised clustering method has a total of n objects for a data set containing n objects 2 There are two possible pairwise relationships, so a large enough number of constraints is required to achieve satisfactory results. In practical applications, obtaining a large number of constraints requires a large amount of human labor and low efficiency, thus providing an optimized Subset grouping semi-supervised fuzzy clustering method

Images