Quantification of Multivariate Categorical Data Considering Typicality of Item
Chi-Hyon Oh, Katsuhiro Honda, Hidetomo Ichihashi · Journal of Advanced Computational Intelligence and Intelligent Informatics · 2007
We propose simultaneously applying homogeneity analysis and fuzzy clustering that simultaneously partitions individuals and items in categorical multivariate datasets. This objective function includes two types of memberships. One is conventional membership representing the degree of membership of each individual in each cluster. The other is an additional parameter that represents typicality of item. A numerical experiment demonstrates that our proposal is useful in quantifying categorical data, taking the typicality of each item into account.