tebak kata dengan algoritma genetika

guestword_ga.js

function randomChar() {
    const chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ ";
    return chars.charAt(Math.floor(Math.random() * chars.length));
}

function randomString(length) {
    return Array.from({ length }, randomChar).join('');
}

function calculateFitness(target, candidate) {
    let score = 0;
    for (let i = 0; i < target.length; i++) {
        if (target[i] === candidate[i]) {
            score++;
        }
    }
    return score / target.length;
}

function crossover(parent1, parent2) {
    let midpoint = Math.floor(Math.random() * parent1.length);
    return parent1.slice(0, midpoint) + parent2.slice(midpoint);
}

function mutate(str, mutationRate = 0.1) {
    return str.split('').map(char => (Math.random() < mutationRate ? randomChar() : char)).join('');
}

function geneticAlgorithm(target, populationSize = 100, mutationRate = 0.1) {
    let population = Array.from({ length: populationSize }, () => randomString(target.length));
    let generation = 0;

    while (true) {
        generation++;

        // Calculate fitness for each candidate
        let scoredPopulation = population.map(candidate => ({
            candidate,
            fitness: calculateFitness(target, candidate)
        }));

        // Sort by fitness (higher first)
        scoredPopulation.sort((a, b) => b.fitness - a.fitness);
        console.log(`scoredPopulation ${JSON.stringify(scoredPopulation)}`);
        // If we found the target, return the result
        if (scoredPopulation[0].candidate === target) {
            console.log(`Found: "${target}" in ${generation} generations!`);
            return;
        }
        // Select the top 50% of the population to be parents
        let parents = scoredPopulation.slice(0, populationSize / 2).map(p => p.candidate);
        console.log(`Parents ${parents}`);
        // Create a new population via crossover and mutation
        population = [];
        while (population.length < populationSize) {
            let parent1 = parents[Math.floor(Math.random() * parents.length)];
            let parent2 = parents[Math.floor(Math.random() * parents.length)];
            let child = crossover(parent1, parent2);
            let childMutate = mutate(child, mutationRate);
            
            console.log(`Generation ${generation}, Parent1 : ${parent1}, Parent2: ${parent2}, Child: ${child}, Mutate: ${childMutate}`);
            population.push(childMutate);
        }

        // console.log(`Generation ${generation}: ${scoredPopulation[0].candidate}, fitness:${scoredPopulation[0].fitness}`);
    }
}

geneticAlgorithm("INFORMATIKA");