ssenet: Semi-supervised Elastic net classification panel

In singha53/ssenet: Semi-supervised implementation of glmnet

Description

build an elastic net classification panel

Usage

ssenet(
  xtrain,
  ytrain,
  alpha,
  lambda = NULL,
  family,
  xtest = NULL,
  ytest = NULL,
  filter = "p.value",
  topranked = 50,
  keepVar = NULL,
  useObsWeights = FALSE,
  max.iter = 100,
  perc.full = 1,
  thr.conf = 0.5
)

Arguments

xtrain	nxp matrix - training dataset
ytrain	factor - response variable
alpha	= 1 (lasso), alpha = 0 (ridge), 0 < alpha < 1 (elastic net penalty)
lambda	= strength of elastic net penalty
family	"binomial" or "multinomial"
xtest	nxp matrx - test dataset
ytest	factor - response variable
filter	= "none" or "p.value"
topranked	= 50 (top number of features to select and build a classifier)
keepVar	- names of specific variable to keep in model
weights	- observational weights; default to 1

singha53/ssenet documentation built on March 17, 2020, 4:41 a.m.