keras_mlp.Rd

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/mlp.R
\name{keras_mlp}
\alias{keras_mlp}
\title{Simple interface to MLP models via keras}
\usage{
keras_mlp(
  x,
  y,
  hidden_units = 5,
  penalty = 0,
  dropout = 0,
  epochs = 20,
  activation = "softmax",
  seeds = sample.int(10^5, size = 3),
  ...
)
}
\arguments{
\item{x}{A data frame or matrix of predictors}

\item{y}{A vector (factor or numeric) or matrix (numeric) of outcome data.}

\item{hidden_units}{An integer for the number of hidden units.}

\item{penalty}{A non-negative real number for the amount of weight decay. Either
this parameter \emph{or} \code{dropout} can specified.}

\item{dropout}{The proportion of parameters to set to zero. Either
this parameter \emph{or} \code{penalty} can specified.}

\item{epochs}{An integer for the number of passes through the data.}

\item{activation}{A character string for the type of activation function between layers.}

\item{seeds}{A vector of three positive integers to control randomness of the
calculations.}

\item{...}{Additional named arguments to pass to \code{keras::compile()} or
\code{keras::fit()}. Arguments will be sorted and passed to either function
internally.}
}
\value{
A \code{keras} model object.
}
\description{
Instead of building a \code{keras} model sequentially, \code{keras_mlp} can be used to
create a feedforward network with a single hidden layer. Regularization is
via either weight decay or dropout.
}
\keyword{internal}