Introduction - Multiobjective Optimization Methodology: A Jumping Gene Approach - page 3

By. transforming. multiple. objectives. into. a. single. itness. value,. as. a. pop-
ulation-based. optimization. algorithm,. the. GA. is. especially. effective. for.
solving. multiobjective. optimization. problems. (MOPs). [25,26].. A. number.
of. multiobjective. evolutionary. algorithms. (MOEAs). have. been. proposed,.
including.the.multiobjective.genetic.algorithm.(MOGA).[25];.niched.Pareto.
genetic.algorithm.2.(NPGA2).[20];.nondominated.sorting.genetic.algorithm.
2. (NSGA2). [17,18];. strength. Pareto. evolutionary. algorithm. 2. (SPEA2). [55];.
Pareto. archived. evolution. strategy. (PAES). [36,37];. microgenetic. algorithm.
(MICROGA).[11,12];.and.so.on.
Except.for.the.PAES,.all.of.these.MOEAs.are.GA.based,.and.they.rely.on.
Pareto.sampling.techniques,.which.are.capable.of.generating.multiple.solu-
tions. in. a. single. simulation. run.. However,. true. Pareto-optimal. solutions.
are. seldom. reached. by. MOEAs.. A. set. of. nondominated. solutions. is. thus.
obtained. instead. [10,16],. and. this. solution. set. is. preferably. as. close. to. the.
true.Pareto-optimal.front.or.reference.front.[10].as.possible..Then,.selecting.
a.compromising.solution.for.a.particular.application.is.the.responsibility.of.
the.decision.maker.
Although. advanced. developments. of. various. MOEAs. together. with.
many. additional. measures. have. been. suggested. (e.g.,. mating. restrictions.
[5,21,22,29,30],. itness. sharing. [27],. and. the. crowding. scheme. [15]),. it. is. still.
not. an. easy. task. to. obtain. a. widespread. nondominated. solution. set,. and.
the. convergence. speed. is. usually. slow.. Thus,. another. biological. genetic.
phenomenon.is.considered.as.a.possible.way.to.improve.search.performance.
for.evolutionary.computation,.which.is.also.the.focus.of.this.topic.
When. evolution. proceeds. generation. by. generation. in. an. MOEA,. the.
genes.in.the.chromosomes.of.the.parent.are.passed.to.the.offspring..This.is.
called.vertical.gene.transmission.(VGT)..In.the.early.period,.it.was.thought.
that. all. genes. in. chromosomes. were. ixed. and. could. be. transferred. only.
through.the.VGT.process..Nevertheless,.this.was.not.true;.biologist.Barbara.
McClintock.first.discovered.that.some.kinds.of.genes.could.“jump”.hori-
zontally. in. the. same. chromosome. or. to. other. chromosomes. in. the. same.
generation.when.there.was.stress.exerted.on.the.chromosomes.[13,24,44]..
Therefore,.these.genes.were.called.jumping.genes.(JGs);.this.phenomenon.
was.termed.horizontal.gene.transmission.(HGT).
The.framework.of.an.MOEA.provides.a.unique.platform.for.the.imitation.
of.the.JG.as.each.fitness.function.can.affect.a.certain.part.(genes).of.the.chro-
mosomes.throughout.the.evolutionary.process..The.striking.findings.of.the.
JG. are. that. it. not. only. offers. better. convergence. but. also. presents. a. wider.
spread.of.nondominated.solutions,.especially.at.both.ends.of.the.true.Pareto-
optimal.front.
This. topic. thus. serves. as. a. key. reference. for. the. introduction. of. the. JG.
in.evolutionary.computation,.covering.the.fundamental.concept,.theoretical.
justification,.simulation.verifications,.and.applications.of.the.JG.