Information Technology Reference
In-Depth Information
Classifiers and Mixed Model for Generated Function
mixed model
cl. 1
cl. 2
cl. 3
data
1.5
1
0.5
0
-0.5
0
0.2
0.4
0.6
0.8
1
Input x
Fig. 8.4.
Classifier models, mixed model and available data for the generated function
Mixed Prediction and Prediction of Classifiers
Fitness and Average Number of Classifiers
6
data
pred +/- 1sd
gen. fn.
cl. 1
cl. 2
max. fitness
avg. fitness
min. fitness
avg. K
150
1.5
5
100
1
4
50
3
0.5
0
2
0
-50
1
-0.5
-100
0
0
0.2
0.4
0.6
0.8
1
0
50
100
150
200
250
Input x
GA iteration
(a)
(b)
Fig. 8.5.
Plots showing the best found model structure for the generated function
using GA model structure search, and fitness and average number of classifiers over
the GA iterations. Plot (a) shows the available data, the model of the classifiers, and
their mixed prediction with 1 standard deviation to either side, and additionally the
mean of the generating function. The matching function parameters of the classifiers are
μ
1
=0
.
09
,σ
1
=0
.
063 and
μ
2
=0
.
81
,σ
2
=0
.
006. Plot (b) shows the maximum, average,
and minimum fitness of the individuals in the population after each GA iteration. The
minimum fitness is usually below the lower edge of the plot. The plot also shows the
average number of classifiers for all individuals in the current population.
of model structure evaluations as the GA. The initial number of classifiers is
after each restart sampled from the binomial distribution
B
(8
,
0
.
5), resulting in
4 classifiers on average.
Search WWH ::
Custom Search