dnnet: Multilayer Perceptron Model for Regression or Classification

In SkadiEye/deepTL: Deep Treatment Learning

Multilayer Perceptron Model for Regression or Classification

Description

Fit a Multilayer Perceptron Model for Regression or Classification

Usage

dnnet(
  train,
  validate = NULL,
  load.param = FALSE,
  initial.param = NULL,
  family = c("gaussin", "binomial", "multinomial", "poisson", "negbin", "coxph",
    "poisson-nonzero", "negbin-nonzero", "zip", "zinb")[1],
  norm.x = TRUE,
  norm.y = ifelse(is.factor(train@y) || (class(train) == "dnnetSurvInput") || (family
    %in% c("poisson", "poisson-nonzero", "zip", "zinb")), FALSE, TRUE),
  activate = "relu",
  n.hidden = c(10, 10),
  learning.rate = ifelse(learning.rate.adaptive %in% c("adam"), 0.001, 0.01),
  l1.reg = 0,
  l2.reg = 0,
  n.batch = 100,
  n.epoch = 100,
  early.stop = ifelse(is.null(validate), FALSE, TRUE),
  early.stop.det = 1000,
  plot = FALSE,
  accel = c("rcpp", "none")[1],
  learning.rate.adaptive = c("constant", "adadelta", "adagrad", "momentum", "adam",
    "amsgrad")[5],
  rho = c(0.9, 0.95, 0.99, 0.999)[ifelse(learning.rate.adaptive == "momentum", 1, 3)],
  epsilon = c(10^-10, 10^-8, 10^-6, 10^-4)[2],
  beta1 = 0.9,
  beta2 = 0.999,
  loss.f = ifelse(is.factor(train@y), "logit", "mse"),
  pathway = FALSE,
  pathway.list = NULL,
  pathway.active = "identity",
  l1.pathway = 0,
  l2.pathway = 0
)

Arguments

train	A dnnetInput or a dnnetSurvInput object, the training set.
validate	A dnnetInput or a dnnetSurvInput object, the validation set, optional.
load.param	Whether initial parameters are loaded into the model.
initial.param	The initial parameters to be loaded.
family	If it's specified as Poisson, a poisson model with log link will be fitted.
norm.x	A boolean variable indicating whether to normalize the input matrix.
norm.y	A boolean variable indicating whether to normalize the response (if continuous).
activate	Activation Function. One of the following, "sigmoid", "tanh", "relu", "prelu", "elu", "celu".
n.hidden	A numeric vector for numbers of nodes for all hidden layers.
learning.rate	Initial learning rate, 0.001 by default; If "adam" is chosen as an adaptive learning rate adjustment method, 0.1 by defalut.
l1.reg	weight for l1 regularization, optional.
l2.reg	weight for l2 regularization, optional.
n.batch	Batch size for batch gradient descent.
n.epoch	Maximum number of epochs.
early.stop	Indicate whether early stop is used (only if there exists a validation set).
early.stop.det	Number of epochs of increasing loss to determine the early stop.
plot	Indicate whether to plot the loss.
accel	"rcpp" to use the Rcpp version and "none" (default) to use the R version for back propagation.
learning.rate.adaptive	Adaptive learning rate adjustment methods, one of the following, "constant", "adam", "amsgrad", "adadelta", "adagrad".
rho	A parameter used in momentum.
epsilon	A parameter used in Adagrad and Adam.
beta1	A parameter used in Adam.
beta2	A parameter used in Adam.
loss.f	Loss function of choice.
pathway	If pathway.
pathway.list	Pathway list.
pathway.active	Activation function for the pathway layer.
l1.pathway	The weight for l1-panelty for the pathway layer.
l2.pathway	The weight for l2-panelty for the pathway layer.

Value

Returns a DnnModelObj object.

See Also

dnnet-class
dnnetInput-class
actF

SkadiEye/deepTL documentation built on Nov. 17, 2022, 1:41 p.m.