Convergence results for an accelerated nonlinear cimmino algorithm
Numerische Mathematik
Acceleration schemes for the method of alternating projections
Journal of Computational and Applied Mathematics
Row projection methods for large nonsymmetric linear systems
SIAM Journal on Scientific and Statistical Computing - Special issue on iterative methods in numerical linear algebra
New iterative methods for linear inequalities
Journal of Optimization Theory and Applications
NE/SQP: a robust algorithm for the nonlinear complementarity problem
Mathematical Programming: Series A and B
A class of methods for solving large convex systems
Operations Research Letters
Projection Support Vector Machine Generators
Machine Learning
Hi-index | 0.00 |
The Convex Inequality Problem (CIP), i.e., find x@?R^n such that Ax=b,g(x)=R^m is a convex function, has been solved by projection algorithms possessing a linear rate of convergence. We propose a projection algorithm that exhibits global and superlinear rate of convergence under reasonable assumptions. Convergence is ensured if the CIP is not empty. A direction of search is found by solving a quadratic programming problem (the projection step). As opposed to previous algorithms no special stepsize procedure is necessary to ensure a superlinear rate of convergence. We suggest a possible application of this algorithm for solving convex constrained Linear Complementarity Problems, i.e., find x@?R^n such thatx=0,Ax+b=0,x,Ax+b=0,g(x)=R^m is a convex function.