Genetic Algorithms Algoritması

1function getFitness(candidate: string, target: string): number {
2  let score = 0;
3  for (let i = 0; i < candidate.length; i++) { if (candidate[i] === target[i]) score++; }
4  return score;
5}
6function crossover(parentA: string, parentB: string): string {
7  const midpoint = Math.floor(Math.random() * parentA.length);
8  return parentA.slice(0, midpoint) + parentB.slice(midpoint);
9}
10function mutate(candidate: string, rate: number, alphabet: string): string {
11  const chars = [...candidate];
12  for (let i = 0; i < chars.length; i++) {
13    if (Math.random() < rate) { chars[i] = alphabet[Math.floor(Math.random() * alphabet.length)]; }
14  }
15  return chars.join('');
16}
17function geneticAlgorithm(target: string, alphabet = '01', size = 100, rate = 0.01): string {
18  let pop: string[] = Array.from({ length: size }, () =>
19    Array.from({ length: target.length }, () => alphabet[Math.floor(Math.random() * alphabet.length)]).join('')
20  );
21  while (true) {
22    pop.sort((a, b) => getFitness(b, target) - getFitness(a, target));
23    if (pop[0] === target) return pop[0];
24    const nextGen: string[] = pop.slice(0, Math.ceil(size * 0.1));
25    while (nextGen.length < size) {
26      const parentA = pop[Math.floor(Math.random() * (size / 2))];
27      const parentB = pop[Math.floor(Math.random() * (size / 2))];
28      let child = crossover(parentA, parentB);
29      child = mutate(child, rate, alphabet);
30      nextGen.push(child);
31    }
32    pop = nextGen;
33  }
34}