@inproceedings{2006004,
abstract = {Search and optimization algorithms are routinely used to find the minimum or maximum solution corresponding to one or more objective functions, subject to satisfying certain constraints. However, monotonicity analysis is a process which, in certain problems having monotonie objectives and constraints, can instantly bring out important properties among decision variables corresponding to optimal solutions, without applying any optimization procedure. Although the approach is appealing, the monotonie restriction on problems is probably the reason for their unpopularity among optimization researchers. In this paper, we suggest a generic two-step evolutionary multi-objective optimization procedure which can bring out important relationships among optimal decision variables and objectives to linear or non-linear optimization problems. Although this "innovization" idea is already put forward by the authors elsewhere, this paper brings out the similarities of the outcome of the proposed innovization task with that of the monotonicity analysis and clearly demonstrates the advantages of the former method in handling generic optimization problems. This paper also demonstrates the ability of multi-objective evolutionary optimization algorithms in finding solutions close to possible theoretical optimal solutions - a matter which has not been attempted adequately in the evolutionary computing literature. {\textcopyright} J. UCS.},
author = {Deb, Kalyanmoy and Srinivasan, Aravind},
booktitle = {Journal of Universal Computer Science},
issn = {0958695X},
keywords = {Design principles,Evolutionary computing,Kuhn-Tucker conditions,Monotonicity analysis,Multi-objective optimization,Pareto-optimal solutions},
number = {7},
pages = {955--970},
title = {{Monotonicity analysis, discovery of design principles, and theoretically accurate evolutionary multi-objective optimization}},
url = {https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.566.1020},
volume = {13},
year = {2007}
}