|
Last change
on this file since 11194 was
6152,
checked in by bfarka, 14 years ago
|
|
added ecj and custom statistics to communicate with the okb services #1441
|
|
File size:
1.3 KB
|
| Line | |
|---|
| 1 | # Copyright 2010 by Sean Luke and George Mason University |
|---|
| 2 | # Licensed under the Academic Free License version 3.0 |
|---|
| 3 | # See the file "LICENSE" for more information |
|---|
| 4 | |
|---|
| 5 | parent.0 = ../../multiobjective/nsga2/nsga2.params
|
|---|
| 6 |
|
|---|
| 7 |
|
|---|
| 8 | # NSGA-II wil use a population size of 100. Its archive will be another 100
|
|---|
| 9 | # on top of that. Since we're not reevaluating elites, only the 100 new
|
|---|
| 10 | # individuals will be bred and evaluated each generation.
|
|---|
| 11 | #
|
|---|
| 12 | # For more discussion about how to get SPEA2 and NSGA-II to have "equivalent"
|
|---|
| 13 | # numbers of evaluations in ECJ, see the README files for the SPEA2 and NSGA-II
|
|---|
| 14 | # packages.
|
|---|
| 15 | #
|
|---|
| 16 | # constants from "Zitzler, E., Deb, K., and Thiele, L., 2000,
|
|---|
| 17 | # Comparison of Multiobjective Evolutionary Algorithms: Empirical Results,
|
|---|
| 18 | # Evolutionary Computation, Vol. 8, No. 2, pp173-195."
|
|---|
| 19 | #
|
|---|
| 20 | pop.subpop.0.size = 100
|
|---|
| 21 | generations = 250
|
|---|
| 22 |
|
|---|
| 23 |
|
|---|
| 24 | # NSGA2's pipeline
|
|---|
| 25 |
|
|---|
| 26 | pop.subpop.0.species.pipe = ec.vector.breed.VectorMutationPipeline
|
|---|
| 27 | pop.subpop.0.species.pipe.likelihood = 1.0
|
|---|
| 28 | pop.subpop.0.species.pipe.source.0 = ec.vector.breed.VectorCrossoverPipeline
|
|---|
| 29 | pop.subpop.0.species.pipe.source.0.likelihood = 0.9
|
|---|
| 30 | pop.subpop.0.species.pipe.source.0.source.0 = ec.select.TournamentSelection
|
|---|
| 31 | pop.subpop.0.species.pipe.source.0.source.1 = same
|
|---|
| 32 | select.tournament.size = 2
|
|---|
| 33 |
|
|---|
Note: See
TracBrowser
for help on using the repository browser.
