cv.alphanorm: Cross-validation for alpha-norm

In yxwang99/alphanorm: alpha-norm regularization

Description

Does k-fold cross-validation for alpha-norm, and return the best lambda and q

Usage

cv.alphanorm(x, y, lambda_Tune = exp(10:-10), q_Tune = c(0.1, 0.5, 0.9),
  intercept = TRUE, nfolds = 5, tol = 1e-07, T = 500, trace = FALSE)

Arguments

x	design matrix
y	response vector
lambda_Tune	user-supplied lambda sequence
q_Tune	user-supplied q sequence
intercept	whether intercept should be in the model, default to be TRUE
nfolds	number of folds , default to be 5
tol	tolerence of convergence condition
T	number of maximum iterations for each coefficient
trace	print the process of alphanorm

Value

An object of S3 class "cv.alphanorm"

lambda	the values of lambda used in the fits in the decreasing order
q	the values of q used in the fits
cvm	The mean cross-validation error, a matrix of length(q)*length(lambda)
lambda.min	value of lambda that gives minimum cvm
q.min	value of q that gives minimum cvm

See Also

alphanorm

yxwang99/alphanorm documentation built on May 23, 2019, 11:34 p.m.