Matrix completion via a low rank factorization model and an Augmented Lagrangean Succesive Overrelaxation Algorithm
Journal: Bulletin of Computational Applied Mathematics (Bull CompAMa) (Vol.2, No. 2)Publication Date: 2014-12-31
Authors : Hugo Lara; Harry Oviedo; Jinjun Yuan;
Page : 21-46
Keywords : matrix completion; alternanting minimization; nonlinear Gauss-Seidel method; nonlinear SOR method; Augmented Lagrange method;
Abstract
The matrix completion problem (MC) has been approximated by using the nuclear norm relaxation. Some algorithms based on this strategy require the computationally expensive singular value decomposition (SVD) at each iteration. One way to avoid SVD calculations is to use alternating methods, which pursue the completion through matrix factorization with a low rank condition. In this work an augmented Lagrangean-type alternating algorithm is proposed. The new algorithm uses duality information to define the iterations, in contrast to the solely primal LMaFit algorithm, which employs a Successive Over Relaxation scheme. The convergence result is studied. Some numerical experiments are given to compare numerical performance of both proposals.
Other Latest Articles
- L(2,1)-Labeling for Subdivisions of Cycle Dominated Graphs
- Uniform Stability and Boundedness of a Kind of Third Order Delay Differential Equations
- A Residual Approach for Balanced Truncation Model Reduction (BTMR) of Compartmental Systems
- At the Mid of Polarity: Rethinking Medium Farms as Solution to Vulnerability for Small-scale Farmers in Sub-Saharan Africa
- Spirituality and Search for Meaning in the Workplace
Last modified: 2018-08-05 10:34:14