Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update Genetic Algorithm implementation in AIMA4e #452

Open
wants to merge 2 commits into
base: AIMA4e
Choose a base branch
from
Open

Update Genetic Algorithm implementation in AIMA4e #452

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c996918
Merge pull request #2 from aimacode/AIMA4e
jsuyash1514 Mar 3, 2019
aed6052
Update Genetic algorithm implementation
jsuyash1514 Mar 4, 2019
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
90 changes: 55 additions & 35 deletions core/src/main/java/aima/core/search/basic/local/GeneticAlgorithm.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,7 +60,7 @@ public class GeneticAlgorithm<A> {
*
* @param population
* a set of individuals
* @param fitnessFn
* @param fitnessFN
* a function that measures the fitness of an individual
* @return the best individual in the specified population, according to the
* specified FITNESS-FN.
Expand All @@ -73,18 +73,8 @@ public Individual<A> geneticAlgorithm(Set<Individual<A>> population, ToDoubleFun
Set<Individual<A>> new_population = new HashSet<Individual<A>>(population.size());
// repeat
do {
// x <- RANDOM-SELECTION(population, FITNESS-FN)
Individual<A> x = randomSelection(population, fitnessFN);
// y <- RANDOM-SELECTION(population, FITNESS-FN)
Individual<A> y = randomSelection(population, fitnessFN);
// child <- REPRODUCE(x, y)
Individual<A> child = reproduce(x, y);
// if (small random probability) then child <- MUTATE(child)
if (isSmallRandomProbabilityOfMutation()) {
child = mutate(child);
}
// add child to new_population
new_population.add(child);
// population <- [MUTATE(RECOMBINE(SELECT(2, population, FITNESS-FN)))]
new_population.add(mutate(recombine(select(2, population, fitnessFN ))));
} // until SIZE(new_population) = SIZE(population)
while (new_population.size() != population.size());
population = new_population;
Expand All @@ -94,13 +84,36 @@ public Individual<A> geneticAlgorithm(Set<Individual<A>> population, ToDoubleFun
return bestIndividual;
}

// function REPRODUCE(x, y) returns an individual
public Individual<A> reproduce(Individual<A> x, Individual<A> y) {
// n <- LENGTH(x);

public Set<Individual<A>> select(int p, Set<Individual<A>> population, ToDoubleFunction<Individual<A>> fitnessFN){
Set<Individual<A>> selection = new HashSet<Individual<A>>(2*p);
// selection <- a uniform random sample of 2*p individuals from population
do{
selection.add(randomSelection(population,fitnessFN));
}
while (selection.size() != 2*p);

// return the top p individuals in selection, ranked by FITNESS-FN
Set<Individual<A>> fittestIndividuals = new HashSet<Individual<A>>(p);
do{
Individual<A> best = selectBestIndividual(selection,fitnessFN);
fittestIndividuals.add(best);
selection.remove(best);
}
while(fittestIndividuals.size() != p);

return fittestIndividuals;
}

public Individual<A> recombine(Set<Individual<A>> parentIndividuals){
Iterator<Individual<A>> parentIt = parentIndividuals.iterator();
Individual<A> x = parentIt.next();
Individual<A> y = parentIt.next();
// n <- LENGTH(X)
int n = x.length();
// c <- random number from 1 to n
// c <- random number from 0 to n
int c = random.nextInt(n);
// return APPEND(SUBSTRING(x, 1, c), SUBSTRING(y, c+1, n))
// return APPEND(x[0:c], y[c:n])
return new Individual<A>(x.substring(0, c), y.substring(c, n));
}

Expand Down Expand Up @@ -154,34 +167,41 @@ public boolean isSmallRandomProbabilityOfMutation() {
}

public Individual<A> mutate(Individual<A> child) {
// We will mutate the child in place.
child.getRepresentation().set(random.nextInt(child.length()), alphabet.get(random.nextInt(alphabet.size())));
// if (small random probability) then child <- MUTATE(child)
if(isSmallRandomProbabilityOfMutation()){
// We will mutate the child in place.
child.getRepresentation().set(random.nextInt(child.length()), alphabet.get(random.nextInt(alphabet.size())));
}
return child;
}

public Individual<A> selectBestIndividualIfReady(Set<Individual<A>> population,
ToDoubleFunction<Individual<A>> fitnessFN) {
Individual<A> best = null;

// until some individual is fit enough, or enough time has elapsed
if (!executionController.isExecuting() || population.stream().anyMatch(fitEnoughPredicate)) {
double bestValue = 0;
Iterator<Individual<A>> popIt = population.iterator();
while (popIt.hasNext()) {
Individual<A> current = popIt.next();
double currentValue = fitnessFN.applyAsDouble(current);
if (best == null || currentValue > bestValue) {
return selectBestIndividual(population,fitnessFN);
}
return null;
}

public Individual<A> selectBestIndividual(Set<Individual<A>> population,
ToDoubleFunction<Individual<A>> fitnessFN) {
Individual<A> best = null;
double bestValue = 0;
Iterator<Individual<A>> popIt = population.iterator();
while (popIt.hasNext()) {
Individual<A> current = popIt.next();
double currentValue = fitnessFN.applyAsDouble(current);
if (best == null || currentValue > bestValue) {
best = current;
bestValue = currentValue;
} else if (currentValue == bestValue) {
// Randomly break ties
if (random.nextBoolean()) {
best = current;
bestValue = currentValue;
} else if (currentValue == bestValue) {
// Randomly break ties
if (random.nextBoolean()) {
best = current;
}
}
}
}

return best;
}
}