The Clustering Analysis Based On Mixed Attribute Information

Clustering is an unsupervised study process, the objective of cluster analysis is to group a set of objects into clusters such that objects within the same cluster have a high degree of similarity, while objects belonging to different clusters have a high degree of dissimilarity. It has been most commonly applied in the economic and the management areas, etc. If the clustering information (feature value of clustering object or similarity matrix or the feature weight) is exactly numerical (crisp) data, there are many literatures on this topic. But for many actual problems, because of the vague and the non-precise of the object's feature, the object's feature values are interval number or triangular fuzzy numbers or linguistic assessment forms. There are some research results to the research of the clustering analysis problems with uncertain information. While the research of the clustering analysis problems with mixed attribute information which contains the definite information or the uncertain information has seldom concern. Therefore, to the research on this kind of question, not only in theory but also in application, there are important significances. This article has carried on the corresponding theory and the method study.At first, this article defines the concept of mixed attribute information and a summary of research results in existence is given. Then the related knowledge about the cluster analysis is introduced. Two clustering methods are proposed based on mixed attribute information. The one is a clustering method based on the judgment category to partial cluster sample, it is to obtain the index weight, and another is a clustering method based on the combined similarity to improve the accuracy. Finally, this paper summarizes the research result and conclusion. Following that the further research needed is pointed out.