Information Technology Reference
In-Depth Information
setosa
; the second is
Iris versicolor
versus NOT
Iris versicolor
; and the
last is
Iris virginica
versus NOT
Iris virginica
.
For all these sub-problems, F = {+, -, *, /} and the set of terminals in-
cludes obviously all the four attributes which were represented by
d
0
-
d
3
,
corresponding, respectively, to sepal length, sepal width, petal length, and
petal width. The 0/1 rounding threshold was set to 0.5 and the fitness was
based on the number of hits and was evaluated by equation (3.8). Thus, for
this problem,
f
max
= 150.
For all the sub-problems, I started with three-genic chromosomes with an
h
= 8 and sub-ETs linked by addition. The first dataset (
Iris setosa
versus
NOT
Iris setosa
) was almost instantaneously classified without errors and I
soon found out that a very simple structure is required to classify correctly
this dataset. The model below perfectly classifies all the irises into
Iris setosa
and NOT
Iris setosa
:
int apsModel(double d[])
{
const double ROUNDING_THRESHOLD = 0.5;
double dblTemp = 0;
dblTemp = (d[1]-d[2]);
return (dblTemp >= ROUNDING_THRESHOLD ? 1 : 0);
(4.11)
}
As you can see, only the difference between the sepal width and the petal
length is relevant to distinguish
Iris setosa
from
Iris versicolor
and
Iris
virginica
.
The classification of the remaining datasets was also extremely accurate,
but on both cases only 149 out of 150 samples were correctly classified. And
in both cases, the sub-models failed only to classify sample 84. The model
below distinguishes
Iris versicolor
from the other two irises:
int apsModel(double d[])
{
const double ROUNDING_THRESHOLD = 0.5;
double dblTemp = 0;
dblTemp = (d[3]*(((d[0]*d[3])-d[1])+((d[1]*d[2])-2])));
dblTemp += (((d[2]-(d[2]+d[2]))-(d[0]/d[0]))/d[3]);
dblTemp += (((d[0]-(d[2]*d[3]))*(d[2]-d[1]))*d[0]);
return (dblTemp >= ROUNDING_THRESHOLD ? 1 : 0);
}
(4.12)
And the next model distinguishes
Iris virginica
from
setosa
and
versicolor
:
