Case B (split training and validation set)
CLASSIFICATION
ALGORITHM
CLASS
SCENARIO
TRAINING
SET
TEST
SET
C
AS
E
B
: s
pl
itti
ng
th
e
K
D
D
Train+
in
tr
ai
ni
ng
an
d
te
st/
vali
dati
on
su
bsets
LOGISTIC
REGRESSION
multi
0,99
0,99
binary
0,97
0,97
4-class
0,99
0,99
DECISION TREE
multi
1,00
1,00
binary
1,00
1,00
4-class
1,00
1,00
K NEAREST
NEIBOURS
multi
0,99
0,99
binary
0,99
0,99
4-class
0,99
0,99
GAUSSIAN
NAÏVE BAYES
multi
0,76
0,76
binary
0,85
0,85
4-class
0,65
0,65
MULTI LAYER
PERCEPTRON
multi
1,00
1,00
binary
1,00
1,00
4-class
1,00
1,00
26