A Survey on Clustering Techniques for Multi-Valued Data Sets

Abstract

The complexity of the attributes in some particular domain is high when compare to the standard domain the reason for this is its internal variation and the structure their representation needs more complex data called multi-valued data which is introduced in this paper Because of this reason it is needed to extend the data examination techniques for example characterization discrimination association analysis classification clustering outlier analysis evaluation analysis to multi-valued data so that we get more exact and consolidated multi-valued data sets We say that multi-valued data analysis is an expansion of the standard data analysis techniques The objects of multi-valued data sets are represented by multi-valued attributes and they contain more than one value for one entry in the data base An example for this type of attribute is languages known this attribute may contain more than one value for the corresponding objects because one person may be known more than one language