/// <inheritdoc/> public IMLMethod Decode(IGenome genome) { NEATPopulation pop = (NEATPopulation)genome.Population; Substrate.Substrate substrate = pop.CurrentSubstrate; return(Decode(pop, substrate, genome)); }
/// <summary> /// Create a sandwich substrate. A sandwich has an input layer connected /// directly to an output layer, both are square. /// </summary> /// <param name="inputEdgeSize">The input edge size.</param> /// <param name="outputEdgeSize">The output edge size.</param> /// <returns>The substrate.</returns> public static Substrate factorSandwichSubstrate(int inputEdgeSize, int outputEdgeSize) { Substrate result = new Substrate(3); double inputTick = 2.0 / inputEdgeSize; double outputTick = 2.0 / inputEdgeSize; double inputOrig = -1.0 + (inputTick / 2.0); double outputOrig = -1.0 + (inputTick / 2.0); // create the input layer for (int row = 0; row < inputEdgeSize; row++) { for (int col = 0; col < inputEdgeSize; col++) { SubstrateNode inputNode = result.CreateInputNode(); inputNode.Location[0] = -1; inputNode.Location[1] = inputOrig + (row * inputTick); inputNode.Location[2] = inputOrig + (col * inputTick); } } // create the output layer (and connect to input layer) for (int orow = 0; orow < outputEdgeSize; orow++) { for (int ocol = 0; ocol < outputEdgeSize; ocol++) { SubstrateNode outputNode = result.CreateOutputNode(); outputNode.Location[0] = 1; outputNode.Location[1] = outputOrig + (orow * outputTick); outputNode.Location[2] = outputOrig + (ocol * outputTick); // link this output node to every input node foreach (SubstrateNode inputNode in result.InputNodes) { result.CreateLink(inputNode, outputNode); } } } return result; }
/// <summary> /// Construct a starting HyperNEAT population. does not generate the /// initial random population of genomes. /// </summary> /// <param name="theSubstrate">The substrate ID.</param> /// <param name="populationSize">The population size.</param> public NEATPopulation(Substrate theSubstrate, int populationSize) : base(populationSize, new FactorHyperNEATGenome()) { SurvivalRate = DefaultSurvivalRate; WeightRange = 5; InitialConnectionDensity = 0.1; RandomNumberFactory = EncogFramework.Instance .RandomFactory.FactorFactory(); CurrentSubstrate = theSubstrate; InputCount = 6; OutputCount = 2; HyperNEATGenome.BuildCPPNActivationFunctions(_activationFunctions); }
/// <inheritdoc/> public IMLMethod Decode(NEATPopulation pop, Substrate.Substrate substrate, IGenome genome) { // obtain the CPPN NEATCODEC neatCodec = new NEATCODEC(); NEATNetwork cppn = (NEATNetwork)neatCodec.Decode(genome); List <NEATLink> linkList = new List <NEATLink>(); IActivationFunction[] afs = new IActivationFunction[substrate.NodeCount]; IActivationFunction af = new ActivationSteepenedSigmoid(); // all activation functions are the same for (int i = 0; i < afs.Length; i++) { afs[i] = af; } double c = this.MaxWeight / (1.0 - this.MinWeight); BasicMLData input = new BasicMLData(cppn.InputCount); // First create all of the non-bias links. foreach (SubstrateLink link in substrate.Links) { SubstrateNode source = link.Source; SubstrateNode target = link.Target; int index = 0; foreach (double d in source.Location) { input.Data[index++] = d; } foreach (double d in target.Location) { input.Data[index++] = d; } IMLData output = cppn.Compute(input); double weight = output[0]; if (Math.Abs(weight) > this.MinWeight) { weight = (Math.Abs(weight) - this.MinWeight) * c * Math.Sign(weight); linkList.Add(new NEATLink(source.ID, target.ID, weight)); } } // now create biased links input.Clear(); int d2 = substrate.Dimensions; IList <SubstrateNode> biasedNodes = substrate.GetBiasedNodes(); foreach (SubstrateNode target in biasedNodes) { for (int i = 0; i < d2; i++) { input.Data[d2 + i] = target.Location[i]; } IMLData output = cppn.Compute(input); double biasWeight = output[1]; if (Math.Abs(biasWeight) > this.MinWeight) { biasWeight = (Math.Abs(biasWeight) - this.MinWeight) * c * Math.Sign(biasWeight); linkList.Add(new NEATLink(0, target.ID, biasWeight)); } } // check for invalid neural network if (linkList.Count == 0) { return(null); } linkList.Sort(); NEATNetwork network = new NEATNetwork(substrate.InputCount, substrate.OutputCount, linkList, afs); network.ActivationCycles = substrate.ActivationCycles; return(network); }
/// <inheritdoc/> public IMLMethod Decode(NEATPopulation pop, Substrate.Substrate substrate, IGenome genome) { // obtain the CPPN NEATCODEC neatCodec = new NEATCODEC(); NEATNetwork cppn = (NEATNetwork)neatCodec.Decode(genome); List<NEATLink> linkList = new List<NEATLink>(); IActivationFunction[] afs = new IActivationFunction[substrate.NodeCount]; IActivationFunction af = new ActivationSteepenedSigmoid(); // all activation functions are the same for (int i = 0; i < afs.Length; i++) { afs[i] = af; } double c = this.MaxWeight / (1.0 - this.MinWeight); BasicMLData input = new BasicMLData(cppn.InputCount); // First create all of the non-bias links. foreach (SubstrateLink link in substrate.Links) { SubstrateNode source = link.Source; SubstrateNode target = link.Target; int index = 0; foreach (double d in source.Location) { input.Data[index++] = d; } foreach (double d in target.Location) { input.Data[index++] = d; } IMLData output = cppn.Compute(input); double weight = output[0]; if (Math.Abs(weight) > this.MinWeight) { weight = (Math.Abs(weight) - this.MinWeight) * c * Math.Sign(weight); linkList.Add(new NEATLink(source.ID, target.ID, weight)); } } // now create biased links input.Clear(); int d2 = substrate.Dimensions; IList<SubstrateNode> biasedNodes = substrate.GetBiasedNodes(); foreach (SubstrateNode target in biasedNodes) { for (int i = 0; i < d2; i++) { input.Data[d2 + i] = target.Location[i]; } IMLData output = cppn.Compute(input); double biasWeight = output[1]; if (Math.Abs(biasWeight) > this.MinWeight) { biasWeight = (Math.Abs(biasWeight) - this.MinWeight) * c * Math.Sign(biasWeight); linkList.Add(new NEATLink(0, target.ID, biasWeight)); } } // check for invalid neural network if (linkList.Count == 0) { return null; } linkList.Sort(); NEATNetwork network = new NEATNetwork(substrate.InputCount, substrate.OutputCount, linkList, afs); network.ActivationCycles = substrate.ActivationCycles; return network; }