nnetBuild: Constructor for nnePtR

In philipmgoddard/nnePtR: Neural Network Classifier

Description

Constructor for nnePtR

Usage

nnetBuild(train_input, train_outcome, nLayers = 1, nUnits = 25,
  lambda = 0.01, seed = 1234, iters = 100, optim_method = "L-BFGS-B",
  trace = FALSE)

Arguments

train_input	data frame or matrix of input features
train_outcome	vector of outcome. should be a factor
nLayers	number of hidden layers
nUnits	number of units in hidden layers. Constant across each hidden layer.
lambda	penalty term
seed	seed for initialising network
iters	number of iterations (passed to optim)
optim_method	opitimisation method (passed to optim)
trace	should report from optim be produced?

philipmgoddard/nnePtR documentation built on May 25, 2019, 5:04 a.m.