MINIMIZING STRUCTURED UNCONSTRAINED FUNCTIONS.
Abstract
For the general nonlinear programming problem (nonlinear constraints, nonlinear objective function) the difficulty of obtaining a solution is controlled by n, the number of problem variables. This is in contrast to linear programming (where nonnegativity requirements are always assumed), where m, the number of nontrivial constraints, determines the size of the problems that can be handled and their speed of solution. This paper outlines in detail modifications to the sequential unconstrained minimization technique (SUMT) for nonlinear programming that expand the size of problems that it can solve to the limitations imposed by the special-case algorithms devised for them. In particular, the special cases of mathematical programming, i.e., linear, separable, and 'factorable,' programming, will be described. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1967
- Accession Number
- AD0661649
Entities
People
- Garth P. Mccormick