DATA MINING AND EXPERIMENTATION
The experiments used 6 classification algorithms in
order to try obtain the highest classification accuracy. The
algorithms are J48, SimpleCart, Kstar, SMO, Naive Bayes,
OneR.
Calculation accuracy used 10 cross-validation method.
Cross-validation method is a statistical algorithms by
dividing data into two segments: one used to learn or train a
model and the the other used to validate the model [15].
The first experiment use all subjects. This is the result of
the experiment in Table I.
From the result, using J48 and OneR give the best
percentage of correctly classified instances 79.61% and
78.66%. Kstar and NaiveBayes give the worst percentage of
correctly classified instances 69.74% and 76.75%.