Graphics Reference
In-Depth Information
(a)
(b)
RNGE (0.1170,0.8930,0.7822)
AllKNN (0.1758,0.8934,0.7831)
(c)
(d)
CPruner (0.8636,0.8972,0.7909)
HMNEI (0.3617,0.8906,0.7787)
(e)
(f)
RMHC (0.9000,0.8972,0.7915)
SSMA (0.9879,0.8964,0.7900)
Fig. 8.5
Data subsets in banana data set (2).
a
RNGE (0
.
1170
,
0
.
8930
,
0
.
7822).
b
AllKNN
(0
.
1758
,
0
.
8934
,
0
.
7831).
(0
.
8636
,
0
.
8972
,
0
.
7909).
(0
.
3617
,
0
.
8906
,
c
CPruner
d
HMNEI
.
.
,
.
,
.
.
,
.
,
.
0
7787).
e
RMHC (0
9000
0
8972
0
7915).
f
SSMA (0
9879
0
8964
0
7900)
The pictures of the subset selected by some PS methods could help to visualize
and understand their way of working and the results obtained in the experimental
study. The reduction rate, the accuracy and kappa values in test data registered in the
experimental study are specified in this order for each one. In original data sets, the
two values indicated correspond to accuracy and kappa with 1NN (Fig.
8.4
a).
