13 ago 2019

Using Machine learning to play Flappy Bird

Program that learns to play Flappy Bird by machine learning (Neuroevolution)

alt tag

NeuroEvolution.js : Utilization

// Initialize
var ne = new Neuroevolution({options});

//Default options values
var options = {
    network:[1, [1], 1],    // Perceptron structure
    population:50,          // Population by generation
    elitism:0.2,            // Best networks kepts unchanged for the next generation (rate)
    randomBehaviour:0.2,    // New random networks for the next generation (rate)
    mutationRate:0.1,       // Mutation rate on the weights of synapses
    mutationRange:0.5,      // Interval of the mutation changes on the synapse weight
    historic:0,             // Latest generations saved
    lowHistoric:false,      // Only save score (not the network)
    scoreSort:-1,           // Sort order (-1 = desc, 1 = asc)
    nbChild:1               // number of child by breeding
}

//Update options at any time
ne.set({options});

// Generate first or next generation
var generation = ne.nextGeneration();

//When an network is over -> save this score
ne.networkScore(generation[x], <score = 0>);

You can see the NeuroEvolution integration in Flappy Bird in Game.js.

Download