Biomedical Engineering Reference
In-Depth Information
2.2.6.3  Three Types of Elitism
Three.types.of.elitism.scheme.are.implemented.in.MICROGA..In.the.first.
type,.nondominated.solutions.found.within.the.internal.cycle.are.stored.
so.valuable.information.obtained.during.the.evolutionary.process.will.not.
be.lost.
For.the.second.type,.nominal.solutions.(i.e.,.the.best.solutions.found.when.
the. nominal. convergence. is. reached). are. added. to. the. replaceable. part. of.
the. population. memory. to. enhance. the. convergence. of. solutions.. It. results.
in.a.higher.probability.of.reaching.the.true.Pareto.front.over.time.using.the.
crossover.and.mutation.operations.
The. last. elitism. scheme. uniformly. picks. a. certain. number. of. solutions.
from.all.the.regions.of.the.Pareto.front.generated.so.far.and.includes.them.in.
the.replaceable.part.of.the.population.memory..Its.purpose.is.to.utilize.the.
best-available.solutions.as.the.starting.point,.and.further.improvements.are.
expected.via.operations.(either.by.getting.closer.to.the.true.Pareto.front.or.by.
getting.a.better.distribution).
2.2.7 ant Colony Optimization
Ant. colony. optimization. (ACO). is. a. metaheuristic. optimization. method.
that. uses. a. model-based. searching. framework. [16,17,23].. It. has. been.
applied. for. solving. many. hard. combinatorial. optimization. problems.
[22,66].
The.principle.of.ACO.can.be.summarized.as.“a.set.of.artificial.ants.mov-
ing.through.states.of.the.problem.corresponding.to.partial.solutions.of.the.
problem.to.solve”.[22],.and.its.lowchart.is.shown.in . Figure 2.12 .. Its.design.
concept. follows. the. foraging. behavior. of. ants.. When. a. population. of. ants.
searches.for.food,.the.ants.are.sent.out.to.explore.the.areas.surrounding.their.
nest.in.a.random.manner..When.an.ant.reaches.the.food,.it.carries.the.food.
back.and.deposits.a.trail.of.chemical.pheromone.that.serves.as.information.
to.guide.the.other.ants.to.the.food.source..The.quantity.of.pheromone.depos-
ited.depends.on.the.quantity.and.quality.of.the.food..After.several.rounds.of.
searching,.the.density.of.the.pheromone.trails.on.the.paths.will.indicate.the.
shortest.way.to.reach.the.food.
In.computation,.the.agents.(i.e.,.the.ants).move.with.a.stochastic.local.deci-
sion.policy..The.agent k .moves.from.state. i .to. j .with.a.probability. p i k ,.which.
depends.on.two.parameters:
.
1.. The.attractiveness.η ij .of.the.move,.indicating.its.a.priori.desirability
.
2.. The.trail.level.τ ij . of.the.move,.indicating.how.proficient.it.has.been.in.
the.past.to.make.the.particular.move,.hence.representing.a.posteriori.
indication.of.the.desirability.of.that.move
Search WWH ::




Custom Search