Institutul de Matematică şi Informatică "Vladimir Andrunachievici"
Universitatea de Stat din MoldovaRO EN
Quasi-Newton Methods for Solving Nonlinear Programming Problems
Authors: V. Moraru
Keywords: Quasi-Newton methods, Constrained Optimization, Superlinear convergence.
– 0.16 Mb
Keywords: Quasi-Newton methods, Constrained Optimization, Superlinear convergence.
Abstract
In the present paper the problem of constrained equality optimization is reduced to sequential solving a series of problems of quadratic programming. The Hessian of the Lagrangian is approximated by a sequence of symmetric positive definite matrices. The matrix approximation is updated at every iteration by a Gram- Schmidt modified algorithm. We establish that methods is locally convergent and the sequence {xk}converges to the solution a two-step superlinear rate.Vasile Moraru, Dept. of Computing Mathematics and Programming, Technical University of Moldova
Bd. Stefan cel Mare, 168, Kishinev, 277012, Moldova
Bd. Stefan cel Mare, 168, Kishinev, 277012, Moldova
Fulltext
– 0.16 Mb
Contents
- The structural robustness of multiprocessor computing system
- On the Family of Conditional Embedded Implicational Dependencies
- Fast software encryption system based on local pseudorandomness
- Quasi-Newton Methods for Solving Nonlinear Programming Problems
- On graphs containing a given graph as median
- Numerical implementation of mechanical quadratures method for solving singular integral equations given on closed smooth contours
- The aproximation of functions in generalized Holder spaces
- Macroeconomic model of national economy development
