An efficient alternating least-squares algorithm to perform multidimensional unfolding

摘要: We consider the problem of least-squares fitting squared distances in unfolding. An alternating procedure is proposed which fixes row or column configuration turn and finds global optimum objective criterion with respect to free parameters, iterating this fashion until convergence reached. A considerable simplification algorithm results, namely that conditional identified by performing a single unidimensional search for each point, irrespective dimensionality unfolding solution.