An analysis of a reordering operator on a GA-hard problem
β Scribed by D. E. Goldberg; C. L. Bridges
- Publisher
- Springer-Verlag
- Year
- 1990
- Tongue
- English
- Weight
- 769 KB
- Volume
- 62
- Category
- Article
- ISSN
- 0340-1200
No coin nor oath required. For personal study only.
β¦ Synopsis
This paper analyzes the performance of a genetic algorithm that combines reproduction, crossover, and a reordering operator. Reordering operators have often been suggested as one way to avoid the coding traps -the combinations of loose linkage and deception among important, lower order schemata-of fixed codings. The analysis confirms this role and suggests directions for further research. blem and its extended schema analysis. This work is then extended by the addition of a simplified model of reordering, the idealized reordering operator (IRO). Numerical and theoretical results demonstrate that the addition of such an operator is sufficient to permit convergence of the combined genetic algorithm for all order-two problems. Ramifications and extensions of this result are also discussed.
π SIMILAR VOLUMES
The following problem is considered: where Ξ» is a spectral parameter. The inverse problem is studied: a subsequence Ξ» n β +β of the sequence of eigenvalues is given and odd f is the unknown quantity. A description of the whole class of solutions of this problem is obtained. In addition, it is prove
It has been proven that if the solution exists to an inhomogeneous biharmonic equation in the plane where the values of the solution, the normal derivative of the solution, and the Laplacian of the solution are prescribed on the boundary, then the domain is a disk. This result has been extended to N