Akademik Veri Yönetim Sistemi
2nd WSEAS International Conference on Multivariate Analysis and Its Application in Science and Engineering, İstanbul, Türkiye, 30 Mayıs - 01 Haziran 2009, ss.72-80
A new and powerful method for matrix decomposition is developed in this work. If is similar to singular Value decomposition and the main idea comes from the univariate approximation of a function, given on a planar grid's nodes. by two variable high dimensional model representation. The proposed method is less iteration dependent than the singular value decomposition and the components arc determined via straight for ward steps containing recursions. It seems to have more capabilities than the singular Value decomposition its an alternative method. An illustrative application is also given.