A combined approach has been applied.A Kohonen self organizing map of a given data set containing only categorical variables is constructed. The output of the first part is used as input to k-means for clustering. It has been observed that it provides better results as compared to directly applying k-means.
A combined approach has been applied.A Kohonen self organizing map of a given data set containing only categorical variables is constructed. The output of the first part is used as input to k-means for clustering. It has been observed that it provides better results as compared to directly applying k-means.
A combined approach has been applied.A Kohonen self organizing map of a given data set containing only categorical variables is constructed. The output of the first part is used as input to k-means for clustering. It has been observed that it provides better results as compared to directly applying k-means.