Generate transfer-learning survmaximin coefficients

Generate transfer-learning survmaximin coefficients

survmaximin(B_source, Sigma_target, delta = 0)

Arguments

B_source	A p (number of variables) * L (number of source sites) numeric matrix.
Sigma_target	Data covariance matrix for the target site.
delta	Default 0; the ridge penalty parameter.

Value

A list with the following components:

beta.est	Estimated beta coefficients for the target site.
weight	A vector containing trained weights for each source site.

Examples

data(B_source); data(Sigma_target)
output <- survmaximin(B_source, Sigma_target, delta=0.5)
beta.maximin <- output$beta.est