Soft clustering with CP matrices

The classical clustering methods such as PCA, ICA, SVM, or most recently, NNMF (the nonnegative matrix factorization method) and its extension, NTF(nonnegative tensor factorization), are fatally based upon an assumption that the number of the groups of the data points is already known. But the fact is there are many cases where we have no idea at all whether there exist some or how many patterns for us to recognize in reality. We introduce a novel clustering method based on CPF, the method of completely positive matrix factorization. An example is supplied to illustrate the implementation of a CPF algorithm.