[ top | up ]
Continuous XOR Benchmark Problem
Usage
mlbench.xor(n, d=2)
Arguments
n
|
number of patterns to create
|
d
|
dimension of the XOR problem
|
Description
The inputs of the XOR problem are uniformly distributed on
the d
-dimensional cube with corners {+-1}. Each pair of
opposite corners form one class, hence the total number of classes is
2^(d-1)
Value
Returns a list with components
x
|
input values
|
class
|
factor vector of length n with target classes
|
Examples
# 2d example
p<-mlbench.xor(300,2)
plot(p$x,col=as.numeric(p$class))
#
# 3d example
p<-mlbench.xor(300,3)
pairs(p$x,col=as.numeric(p$class))