Abstract
A steepest ascent family of algorithms suitable for the direct solution of continuous variable unconstrained nonconical multiple objective programming problems is introduced. Nonconical multiple objective problems, unlike standard (conical) vector optimization problems, cannot be easily solved by examining related single objective problems. The concept of a direction of steepest ascent is generalized to the multiple objective context and the question of algorithmic convergence is treated. A computational example involving a nonconical unanimity order is given.
Original language | English (US) |
---|---|
Pages (from-to) | 188-221 |
Number of pages | 34 |
Journal | Journal of Mathematical Analysis and Applications |
Volume | 100 |
Issue number | 1 |
DOIs | |
State | Published - Apr 30 1984 |
Funding
This research was partially supported by National Science Foundation Grant ECS-8105965 to Northwestern University. I wish to thank Martin Czigler, who performed some of the computational work of Section 6.
ASJC Scopus subject areas
- Analysis
- Applied Mathematics