Uses of Interface
de.jungblut.math.activation.ActivationFunction
-
Packages that use ActivationFunction Package Description de.jungblut.classification.nn de.jungblut.math.activation -
-
Uses of ActivationFunction in de.jungblut.classification.nn
Fields in de.jungblut.classification.nn declared as ActivationFunction Modifier and Type Field Description ActivationFunction[]MultilayerPerceptronCostFunction.NetworkConfiguration. activationsMethods in de.jungblut.classification.nn that return ActivationFunction Modifier and Type Method Description ActivationFunction[]MultilayerPerceptron. getActivations()Methods in de.jungblut.classification.nn with parameters of type ActivationFunction Modifier and Type Method Description static MultilayerPerceptron.MultilayerPerceptronBuilderMultilayerPerceptron.MultilayerPerceptronBuilder. create(int[] layer, ActivationFunction[] activations, LossFunction errorFunction, Minimizer minimizer, int maxIteration)Creates a new TrainingConfiguration with the mandatory configurations of the activation functions, the to be used minimizer and the maximum iterations.static RBM.RBMBuilderRBM.RBMBuilder. create(ActivationFunction activation, int... layer)Creates a newRBM.RBMBuilderfrom an activation function and layersizes.static RBMRBM. single(int numHiddenNodes, ActivationFunction func)static RBMRBM. singleGPU(int numHiddenNodes, ActivationFunction func)static RBMRBM. stacked(ActivationFunction func, int... numHiddenNodes)Creates a new stacked RBM with sigmoid activation and with the given number of hidden nodes in each stacked layer.static RBMRBM. stackedGPU(ActivationFunction func, int... numHiddenNodes)Creates a new stacked RBM with sigmoid activation and with the given number of hidden nodes in each stacked layer.Constructors in de.jungblut.classification.nn with parameters of type ActivationFunction Constructor Description RBMCostFunction(de.jungblut.math.DoubleVector[] currentTrainingSet, int batchSize, int numThreads, int numHiddenUnits, ActivationFunction activationFunction, TrainingType type, double lambda, long seed, boolean stochastic) -
Uses of ActivationFunction in de.jungblut.math.activation
Classes in de.jungblut.math.activation that implement ActivationFunction Modifier and Type Class Description classAbstractActivationFunctionImplements the boiler plate code for applying functions on container classes like vectors and matrices by applying the function on every element.classElliotActivationFunctionImplementation of the elliot activation function.classLinearActivationFunctionLinear activation function.classLogActivationFunctionLog activation function, guarded against NaN and infinity edge cases.classReluActivationFunctionRectified linear units implementation.classSigmoidActivationFunctionImplementation of the sigmoid function.classSoftMaxActivationFunctionSoftmax activation that only works on vectors, because it needs to sum and divide the probabilities.classSoftplusReluActivationFunctionSmoothed approximation to aReluActivationFunction.classStepActivationFunctionClassic perceptron-like step function.classTanhActivationFunctionImplementation of the Tanh activation based onFastMath.Methods in de.jungblut.math.activation that return ActivationFunction Modifier and Type Method Description ActivationFunctionActivationFunctionSelector. get()
-