Abstract
In this paper, a neurodynamic approach is proposed for solving multiobjective linear programming problems. Multiple objectives are firstly scalarized using a weighted sum technique. Recurrent neural networks are then adopted to generate Pareto-optimal solutions. To diversify the solutions along Pareto fronts, particle swarm optimization is used to optimize the weights of the scalarized objective function. Numerical results are presented to illustrate the effectiveness of the proposed approaches.
Author supplied keywords
Cite
CITATION STYLE
Leung, M. F., & Wang, J. (2018). A neurodynamic approach to multiobjective linear programming. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10878 LNCS, pp. 11–18). Springer Verlag. https://doi.org/10.1007/978-3-319-92537-0_2
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
