elnet_coord: Elastic net regression

In sophiaycl/6520HW2: STSCI6520HW

Usage

elnet_coord(X, Y, a = 0.5, lambda = 1, tol = 10^(-8),
  iterMax = 100, beta0 = rep(0, dim(X)[2]), GLMNET = FALSE)

Arguments

X	Covariate matrix
Y	Response vector
a	[0,1] specifying the relative weight between L1 penalty and L2 penalty
lambda	Non-negative parameter
tol	Tolerance error
iterMax	Maximum iteration times
beta0	Initial guess for coefficients beta
GLMNET	use 'glmnet' or not. Solve the elastic net regression problem. And the objective function to be minimized here is L = ||Y - X beta||_2^2 + lambda ((1 - a)||beta||_2^2 + a ||beta||_1). elnet_coord(matrix(rnorm(300), nrow = 50), runif(50), beta0 = rep(1,4)) glmnet

sophiaycl/6520HW2 documentation built on June 6, 2019, 8:36 p.m.