UPEI: Universal Prediction Expected Improvement UP-EI criteria

In malekbs/UP: What the package does (short line)

Description

Compute the universal preidction expected improvement

Usage

UPEI(x, model, plugin = NULL, alpha = 0.01, up_method = "Empirical",
  envir = NULL)

Arguments

x	points of the design space in which the criterion will be computed
model	the surrogate model
plugin	the current minimum value
alpha	a parameter for exploration term
up_method	type of computation, default "Empirical". Other options: "UP_reg" regularized emprirical or "assume_gauss" regularize UP as Gaussian ditribution.
envir	envirement variable

Value

the value of the UP expected improvement

Examples

#' library(UP)
d            <- 2
n            <- 16
X            <- expand.grid(x1=s <- seq(0,1, length=5), x2=s)
Xtest        <- expand.grid(x1=seq(0,1,length=6), x2=seq(0,1,length=6))
Y            <- apply(X, 1, branin)
upsm         <- UPSM$new(sm= krigingsm$new(), UP=UPClass$new(X,Y,Scale =TRUE))
crit         <- UPEI(x= t(Xtest), model=upsm,plugin=min(Y)) 
print(max(crit))

malekbs/UP documentation built on May 14, 2019, 8:05 a.m.