e.scal

0th

Percentile

Function to calculate e- or softmax scaled membership values

Calculates the e- or softmax scaled membership values of an argmax based classification rule.

Keywords
classif
Usage
e.scal(x, k = 1, tc = NULL)
Arguments
x
matrix of membership values
k
parameter for e-scaling (1 for softmax)
tc
vector of true classes (required if k has to be optimized)
Details

For any membership vector y $\exp(y\cdot k) / \sum\exp(y\cdot k)$ is calculated. If k=1, the classical softmax scaling is used. If the true classes are given, k is optimized so that the apparent error rate is minimized.

Value

  • A list containing elements
  • svScaled values
  • kOptimal k

concept

  • Softmax scaling
  • E scaling
  • Classification
  • Membership values

References

Garczarek, Ursula Maria (2002): Classification rules in standardized partition spaces. Dissertation, University of Dortmund. URL http://hdl.handle.net/2003/2789

Aliases
  • e.scal
Examples
library(MASS)
data(iris)
ldaobj <- lda(Species ~ ., data = iris)
ldapred <- predict(ldaobj)$posterior
e.scal(ldapred)
e.scal(ldapred, tc = iris$Species)
Documentation reproduced from package klaR, version 0.6-11, License: GPL-2

Community examples

Looks like there are no examples yet.