An Adaptive Fuzzy Min-Max Neural Network Classifier Based on Principle Component Analysis and Adaptive Genetic Algorithm

Table 7

Comparison with other traditional classifiers.

Technique	Misclassification
Bayes classifier^1	2
k-nearest neighborhood^1	4
Fuzzy k-nn^2	4
Fisher ratios^1	3
Ho-kashyap^1	2
Perceptron^3	3
Fuzzy perceptron^3	2
FMNN^1	2
GFMN^1	1/0
GFMN^3	0
FMCN^1	0
FMCN^3	0
AFMN^1	0
AFMN^3	0

1Training set is of 75 data points (25 from each class) and test set consists of remaining data points.
²Training data is of 36 data points (12 from each class) and test set consists of 36 data points; results are then scaled up for 150 points.
³Training and testing data are the same.