The purpose of this section is to see if it is effective to
use simple random sampling with very small sample size.
To this end, we test the ensemble classifier on 5Statlog data
sets: Satimage, Segment, Shuttle, Australian, and DNA. For
data description, please see Table 3. The reason to use these
5 Statlog data sets is because Ankerst used them as benchmark
in his PBC system [2]. The experimental results in
Table 4 show that even with very small sample size, the
ensemble classification accuracy is comparable with PBC
using complete training data set. And the average tree size
of ensemble classifier using small sample is significantly
smaller than PBC. For an example of ensemble classifier visualization
on Satimage (satellite image) data, please refer
to Fig.7