Examples of de.jungblut.math.minimize.GradientDescent

de.jungblut.math.minimize.GradientDescent
Gradient descent implementation with some neat features like momentum, divergence detection, delta breaks and bold driver and scheduled annealing adaptive learning rates. For more sophisticated configuration use the {@link GradientDescentBuilder}. @author thomas.jungblut

   * @param alpha the learning rate for gradient descent.
   * @param numIterations how many iterations of training have to be done. (if
   *          converged before, it will stop training)
   */
  public void train(DoubleVector[] trainingSet, double alpha, int numIterations) {
    train(trainingSet, new GradientDescent(alpha, 0d), numIterations);
  }

View Full Code Here

    // test the linear regression case
    // use a gradient descent with very small learning rate
    MultilayerPerceptron mlp = MultilayerPerceptron.MultilayerPerceptronBuilder
        .create(new int[] { 2, 1 },
            new ActivationFunction[] { LINEAR.get(), LINEAR.get() },
            new SquaredMeanErrorFunction(), new GradientDescent(1e-8, 6e-5),
            10000).verbose(false).build();


    // sample a line of points
    Tuple<DoubleVector[], DoubleVector[]> sample = sampleLinear();

View Full Code Here

TOP

Related Classes of de.jungblut.math.minimize.GradientDescent

de.jungblut.classification.nn.MultiLayerPerceptronTest

de.jungblut.classification.nn.RBM

de.jungblut.math.DoubleVector

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.