af.nnet: Function af.nnet

In AndreasFischer1985/quantqual: Software package for analyzing quantitative and qualitative data.

Function af.nnet

Description

Fits multiple initializations of nnet to scaled data and returns best nnet-object.

Usage

af.nnet(data.train, output = NULL, rep.nnet = 10, seed = 0,
  plot = F, size = NULL, decay = 0, linout = T, trace = F,
  maxit = 1000, preProc = c("center", "scale"),
  control = caret::trainControl(method = "boot", number = 10), ...)

Arguments

data.train	data.frame containing training data.
output	index or name of criterion in training data.
size	numeric value determining the number of hidden units.
decay	numeric value determining the decay parameter.

Details

Fits multiple initializations of nnet to scaled data and returns best nnet-object. If size or decay are NULL, 10-fold cross validation is applied to find an optimal set of parameters

AndreasFischer1985/quantqual documentation built on June 20, 2022, 4:55 p.m.