cv.psiITRVS: Cross validation for psiLearn learning with variable...

In mylzwq/psi-learning-for-ITR: An Implementation of psiLearn on Individualized Treatment Rule

Description

Return the psi-learning models with best tuning parameters.

Usage

cv.psiITRVS(X,A,R,m=5,kernel='linear',kappa.ratio=0.01,kappa.max=1,nkappa=10,tau=0.1,lambda.ratio=0.1,lambda.max=2,nlam=10,tau2=0.2,maxit=100,tol=1e-5,res=FALSE)

Arguments

X	n by p input matrix.
A	a vector of n entries coded 1 and -1 for the treatment assignments.
R	a vector of outcome variable, larger is more desirable.
m	m-folds cross validation
kernel	the kernel used in the decision, here we only use the linear kernel.
kappa.ratio	the ratio between the max kappa and the min kappa which controls the complexity of the decision function.
kappa.max	max kappa in the tunning parameter seq.
nkappa	num of kappa for grid search.
tau	tuning parameter for the ridge penalty in psi-Learning.
lambda.ratio	the ratio between the max lambda and the min lambda which controls the complexity of the decision function in the TLP penalty.
lambda.max	max lambda in the tunning parameter seq
nlam	num of kappa for grid search
tau2	tunning parameter to control the margin used in TLP penalty,default is 0.2
maxit	number of max iteration used in psiITR
tol	tolerance used in psiITR
res	Whether to estimate the residual as the outcome for interaction effect, default is FALSE

Value

It returns the estimated coefficients in the decision funcion after cross validation

w	the coefficent for the decision function.
bias	the intercept in both the linear case and the kernel case.
sigma	if the kernel is rbf then the optimal sigma is returned

Author(s)

MingyangLiu <[e-mail]>

Examples

          n=100;p=5
          X=replicate(p,runif(n, min = -1, max = 1))
          A=2*rbinom(n, 1, 0.5)-1
          T=cbind(rep(1,n,1),X)%*%c(1,2,1,0.5,rep(0,1,p-3))
          T0=(cbind(rep(1,n,1),X)%*%c(0.54,-1.8,-1.8,rep(0,1,p-2)))*A
          R=as.vector(rnorm(n,mean=0,sd=1)+T+T0)
          cv_psi_LinearVS<-cv.psiITRVS(X,A,R,m=5,kernel='linear',kappa.ratio=0.01,kappa.max=1,nkappa=10,tau=0.1,lambda.ratio=0.1,lambda.max=2,nlam=10,tau2=0.2,maxit=100,tol=1e-5,res=FALSE)

mylzwq/psi-learning-for-ITR documentation built on May 15, 2019, 1:18 p.m.