FeedForwardNeuralNetworks

FeedForwardNeuralNetwork.js

Create a new Feedforward neural network model.

new FeedForwardNeuralNetworks(options: object?)
Parameters
options (object?)
Name Description
options.hiddenLayers Array (default [10]) Array that contains the sizes of the hidden layers.
options.iterations number (default 50) Number of iterations at the training step.
options.learningRate number (default 0.01) Learning rate of the neural net (also known as epsilon).
options.regularization number (default 0.01) Regularization parameter af the neural net.
options.activation string (default 'tanh') activation function to be used. (options: 'tanh'(default), 'identity', 'logistic', 'arctan', 'softsign', 'relu', 'softplus', 'bent', 'sinusoid', 'sinc', 'gaussian'). (single-parametric options: 'parametric-relu', 'exponential-relu', 'soft-exponential').
options.activationParam number (default 1) if the selected activation function needs a parameter.
Static Members
load(model)
Instance Members
predict(features)
toJSON()
train(features, labels)