Solving linearly constrained nonlinear programming problems by Newton's method

Author(s)

Ghotb, Fatemeh

Available versions

Abstract

In this paper matrix techniques are developed to implement a modified Newton method to a linearly constrained nonlinear optimisation problem with a twice differentiable and factorable objective function. The problem first is reduced to an unconstrained problem, then the generalised inverse of the positive part of the Hessian matrix is used to generate the direction of search.

Publication year

1986

Publication type

Working paper

Source

Faculty of Business staff papers, no. 30

Publisher

Swinburne Institute of Technology

Copyright

Copyright © 1986 Fatemeh Ghobt.

Details