public void Learn(NeuralLayer fromLayer) { for (int i = 0; i < Neurons.Count; i++) { var valueDerivatives = Activate.SoftMaxDer(Neurons[i].OutputDendrites.Select(d => d.Signal.Value).ToArray()); for (int d = 0; d < Neurons[i].OutputDendrites.Count; d++) { var currentDendrite = Neurons[i].OutputDendrites[d]; foreach (var fromNeuron in fromLayer.Neurons) { foreach (var fromDendrite in fromNeuron.InputDendrites) { currentDendrite.Gamma += fromDendrite.Gamma * currentDendrite.Weight; } } currentDendrite.Gamma *= ActivationType == ActivationType.SoftMax ? valueDerivatives[d] : Activate.GetActivationDer(ActivationType, currentDendrite.Signal.Value); currentDendrite.UpdateWeightAndBias(); } //foreach (var fromNeuron in fromLayer.Neurons) //{ // Neurons[i].Gamma += fromNeuron.Gamma * Neurons[i].Weight; //} //var averageDendriteOutput = Neurons[i].OutputDendrites.Average(d => d.Signal.Value); //Neurons[i].Gamma *= Activate.GetActivationDer(ActivationType, averageDendriteOutput); //Neurons[i].UpdateWeightAndBias(); } }
public void Learn(double[] expectedOutput) { for (int i = 0; i < Neurons.Count; i++) { if (ActivationType == ActivationType.SoftMax) { var valueDerivatives = Activate.SoftMaxDer(Neurons[i].OutputDendrites.Select(d => d.Signal.Value).ToArray()); for (int d = 0; d < Neurons[i].OutputDendrites.Count; d++) { Neurons[i].OutputDendrites[d].Gamma = (Neurons[i].OutputDendrites[d].Signal.Value - expectedOutput[d]) * valueDerivatives[d]; Neurons[i].OutputDendrites[d].UpdateWeightAndBias(); } } else { //foreach (var dendrite in Neurons[i].OutputDendrites) //{ // Neurons[i].Gamma += (dendrite.Signal.Value - expectedOutput[i]) * Activate.GetActivationDer(ActivationType, dendrite.Signal.Value); //} for (int d = 0; d < Neurons[i].OutputDendrites.Count; d++) { var signal = Neurons[i].OutputDendrites[d].Signal; var value = (signal.Value - expectedOutput[d]) * Activate.GetActivationDer(ActivationType, signal.Value); Neurons[i].OutputDendrites[d].Gamma = value; Neurons[i].OutputDendrites[d].UpdateWeightAndBias(); } } //var averageDendriteOutput = Neurons[i].OutputDendrites.Average(d => d.Signal.Value); //Neurons[i].Gamma = // (averageDendriteOutput - expectedOutput[i]) * // Activate.GetActivationDer(ActivationType, averageDendriteOutput); //Neurons[i].UpdateWeightAndBias(); } }