We usually consider the method of classification and clustering for analysis of homogeneous data. But we know that the criterion for homogeneous item is to get from the data. We consider the Multidimensional scaling techniques for a set of observed si...
We usually consider the method of classification and clustering for analysis of homogeneous data. But we know that the criterion for homogeneous item is to get from the data. We consider the Multidimensional scaling techniques for a set of observed similarities between every pair of items to find a representation of the items in few dimensions such that the interitem proximities nearly match the original similarities. Therefore, we propose the method for the proper decision of similarities and homogeneous items in this paper.