ELMSurv (version 0.6)

cv.ELMCoxBAR: Cross-Validation for ELMCoxBAR

Description

Cross-Validation for ELMCoxBAR to tune kernel parameters Based on log Likelihood

Usage

cv.ELMCoxBAR(x, y, Kernel_type = "lin_kernel", Kernel_para = c(2, 1),
  penality = 0, maxiter = 5, nfolds = 2, ...)

Arguments

x

The covariates(predictor variables) of training data.

y

Survival time and censored status of training data. Must be a Surv survival object

Kernel_type

Type of kernel matrix. Currently four options avaibable. "RBF_kernel",a RBF kernel;"lin_kernel" , a linear kernel;poly_kernel ,a polynomial kernel;sigmoid_kernel, a sigmoid kernel. Default is "lin_kernel".

Kernel_para

Parameters for different types of kernels. A single value for RBF and linear kernels. A vector for polynomial and sigmoid kernels and progam stops if only a single value is supplied. However, if the vector of values is supplied in the cases of RBF and liner kernels, only the first value will be used. Default is a vector value "c(2,1)".

penality

Currently, penality is defaulted to 0 to train an ELMCoxBAR model.

maxiter

Maximum values of iterations to update the CoxBAR estimator. Default is 5.

nfolds

Number of folds in cross validation.

...

Additional arguments for glmnet.

Value

Object of class ELMCoxBAR with elements

elmcox	A glmnet type model. See glmnet for details.
trainx	Training data covariates.
kerneltype	Type of kernel matrix used in training. kerneltype=1,a RBF kernel;kerneltype=2 , a linear kernel;kerneltype=3 ,a polynomial kernel;kerneltype=4, a sigmoid kernel.
Kernel_para	Optimal Parameters returned by cross validation. A single value for kerneltype=1 or 2. A vector for kerneltype=3 or 4.

References

• Wang, H, Li, G. Extreme learning machine Cox model for high-dimensional survival analysis. Statistics in Medicine. 2019; 38:2139-2156.

Examples

# NOT RUN {
set.seed(123)
require(ELMSurv)
require(survival)
#Lung DATA
data(lung)
lung=na.omit(lung)
lung[,3]=lung[,3]-1
n=dim(lung)[1]
L=sample(1:n,ceiling(n*0.5))
trset<-lung[L,]
teset<-lung[-L,]
rii=c(2,3)
# Default with lin_kernel 
elmsurvmodel=cv.ELMCoxBAR(x=trset[,-rii],y=Surv(trset[,rii[1]], trset[,rii[2]]))
#The predicted linear predictor
bestpara=elmsurvmodel$Kernel_para
# }