Examples of eas.users.lukas.neuroCEGPM.simpleNeural.activationFunction.ActivationFunction

Package eas.users.lukas.neuroCEGPM.simpleNeural.activationFunction

Examples of eas.users.lukas.neuroCEGPM.simpleNeural.activationFunction.ActivationFunction

eas.users.lukas.neuroCEGPM.simpleNeural.activationFunction.ActivationFunction
Interface modeling the activation function Phi: R=> R of a neuron. Do not confuse with TransferFunction which models the computation of the weighted input of a neuron. (In literature the term transfer function is often referring to the activation function as well.) IMPORTANT: Activation functions are constrained to return values from [0, 1]. @author Lukas König

        if (endTime - startTime > longTime) {
            GlobalVariables.getPrematureParameters().logWarning("Very long matrix multiplication in propagate: " + (endTime - startTime));
        }


        for (int i = 0; i < this.neuronVector.getRowCount(); i++) {
            ActivationFunction act = this.getActivationFunction(i);
            this.neuronVector.set(0, i, act.activationPhi(this.neuronVector.get(0, i)));
        }
        
        if (normalize) {
            this.normalizeNeuronVector(maxNeuronValue);
        }

View Full Code Here

    public void propagateReverse(boolean normalize) {
        if (this.inverseWeightMatrix == null) {
            this.createInverseWeightMatrix();
        }
        for (int i = 0; i < this.neuronVector.getRowCount(); i++) {
            ActivationFunction act = this.getActivationFunction(i);
            
            try {
                this.neuronVector.set(0, i, act.inverseActivationPhi(this.neuronVector.get(0, i)));
            } catch (InversionNotSupportedException e) {}
        }
        
        long startTime = System.currentTimeMillis();
        this.neuronVector.mult(this.inverseWeightMatrix);

View Full Code Here

    private void createInverseWeightMatrix() {
        this.inverseWeightMatrix = this.weightMatrix.inverse();
    }
    
    public ActivationFunction getActivationFunction(int neuronID) {
        ActivationFunction a = this.activationFunctions.get(neuronID);
        if (a == null) {
            return this.standardActFct;
        } else {
            return a;
        }

View Full Code Here

     * @param sequence  The sequence to translate.
     * 
     * @return  The translation.
     */
    public NeuroController translateGenomeToController(List<BigDecimal> sequence, boolean normalize) {
        ActivationFunction func = this.getStandardActFct();
        int scale = MatrixBigDecimal.GLOBAL_SCALE;
        MatrixBigDecimal.GLOBAL_SCALE = 10;
        this.setStandardActFct(new ActivationFunctionSQRT());
        
//        System.out.println();

View Full Code Here

TOP

Related Classes of eas.users.lukas.neuroCEGPM.simpleNeural.activationFunction.ActivationFunction

eas.users.lukas.neuroCEGPM.simpleNeural.NeuralNetwork

eas.users.lukas.neuroCEGPM.simpleNeural.NeuroTranslator

org.jfree.chart.axis.NumberAxis

org.jfree.chart.JFreeChart

org.jfree.chart.plot.XYPlot

org.jfree.chart.renderer.xy.XYLineAndShapeRenderer

org.jfree.chart.title.TextTitle

org.jfree.data.xy.XYSeries

org.jfree.data.xy.XYSeriesCollection

java.awt.BasicStroke

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.