wfgTrafos: WFG Transformations

Description Usage Arguments Value

Description

tNone is a transformation that changes nothing. It is used to move the entries cursor along to change later entries.
tPoly is the polynomial bias transformation.
tFlat creates a region in search space in which all points have the same objective values.
tParam is the parameter-dependent transformation.
tLinear creates a linear shift of the true optimum.
tDecept creates regions in the search space that have a sub-optimal value but larger area.
tMulti creates many local optima.
tSum creates a dependence between different search-space entries.
tNonsep creates a dependence between objectives.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
wfgTrafos()

tNone(y)

tPoly(y, alpha = 0.02)

tFlat(y, value = 0.8, from = 0.75, to = 0.85)

tParam(y, y.prime = NA, factor = 0.98/49.98, starter = 0.02, ender = 50)

tLinear(y, zero.loc = 0.35)

tDecept(y, opti.loc = 0.35, aperture = 0.001, deceptive.value = 0.05)

tMulti(y, num.minima = 30, hill.size = 95, opti.loc = 0.35)

tSum(y, i, k, M, from = NA, to = NA, weights = NA)

tNonsep(y, degree = NA)

Arguments

y

The value of the search space entry to which to apply this to.

alpha

tPoly: alpha>1 biases toward 0, <1 biases toward 1.

value

tFlat: The value of the flat region. The adjacent regions interpolate to this value.

from

tFlat: The region of the search space that is flat, same for every dimension.
tSum: The entries which are made dependent on eachother

to

see from

y.prime

tParam: Is set by the system, it does not occur in the parameters the user writes in the specification

factor

tParam: Determines the slope of the ramp.

starter

tParam: The influence region.

ender

see starter

zero.loc

tLinear: The location of the true optimum.

opti.loc

tDecept and tMulti: The location of the true optimum.

aperture

tDecept: The size of the opening around the true optimum.

deceptive.value

tDecept: The value of the sub-optimal areas.

num.minima

tMulti: The number of the local optima.

hill.size

tMulti: The size of the hills between the local optima.

i

tSum: i, k, M are set by the system, they do not occur in the parameters the user writes in the specification

k

see i

M

see i

weights

tSum: Optional for the weighted sum.

degree

tNonsep: Degree of nonseparability.

Value

The modified value.


tudob/wfg documentation built on June 1, 2019, 2:54 a.m.