Matrix completion via a low rank factorization model and an Augmented Lagrangean Succesive Overrelaxation Algorithm
Journal Title: Bulletin of Computational Applied Mathematics (Bull CompAMa) - Year 2014, Vol 2, Issue 2
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.
Authors and Affiliations
Hugo Lara, Harry Oviedo, Jinjun Yuan
Keywords
Related Articles
- EP ID EP245119
- DOI -
- Views 83
- Downloads 0
Matrix completion via a low rank factorization model and an Augmented Lagrangean Succesive Overrelaxation Algorithm
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 iter...
Some Convergence Strategies for the Alternating Generalized Projection Method
In this paper we extend the application of the alternating projection algorithm to solve the problem of finding a point in the intersection of $n$ sets ($n\geq2$), which are not all of them convex sets. Here we term such...
A new flexible extension of the generalized half-normal lifetime model with characterizations and regression modeling
In this paper, we introduced a new flexible extension of the Generalized Half-Normal lifetime model as well as a new log-location regression model based on the proposed model. Some useful characterization results are pre...
Solving the KPI wave equation with a moving adaptive FEM grid
The Kadomtsev-Petviashvili I (KPI) equation is the difficult nonlinear wave equation $U_{xt} + 6U_x^2 + 6UU_{xx} + U_{xxxx} = 3U_{yy}.$ We solve this equation using PDE2D (www.pde2d.com) with initial conditions consisti...
Algebraic structures in generalized Clifford analysis and applications to boundary value problems
The present article has a threefold purpose: First it is a survey of the algebraic structures of generalized Clifford-type algebras and shows the main results of the corresponding Clifford-type analysis and its applicati...