Nonnegative Matrix Factorization With Nonlinier Constraints
MetadataShow full item record
Processing of data with very large dimension has been a hot topic in recent decades. Various techniques have been proposed in order to execute the desire information or structure. Non-negative Matrix Factorization based on non-negatives data has become one of the popular methods for shrinking dimension. The main strength of this method is non-negative object, the object model by a combination of some basic non-negative parts so as to provide a physical interpretation of the object construction. NMF methods include the use of text mining, pattern recognition and bioinformatics. Mathematical formulation for NMF did not appear as convex optimization problem and various types of algorithms have been proposed to solve the problem. Framework for Alternative Nonnegative Least Square (ANLS) are coordinates of the block formulation approaches that have been proven reliable theoretically and empirically efficient. This research proposes a new a new algorithm to solve NMF based on the framework ANLS. This algorithm is put forward main pivot methods to least squares problem with non – negative constraints which can overcome the limitations of active set method. The proposed algorithm explores the reduce gradient method is a method that efficiently block the main pivot in context of NMF. ANLS convergence properties of frameworks also owned by these algorithms can be developed for the formulations and NMF with other constraints.