Case A (separate training and test sets)
CLASSIFICATION
ALGORITHM
CLASS
SCENARIO
TRAINING
SET
TEST
SET
C
AS
E
A
: u
si
ng
th
e
K
D
D
Train+
an
d
K
D
D
Test+
as
t
rainin
g
an
d
te
st
set
s
LOGISTIC
REGRESSION
multi
0,99
0,70
binary
0,97
0,75
4-class
0,99
0,76
DECISION TREE
multi
1,00
0,71
binary
1,00
0,79
4-class
1,00
0,76
K NEAREST
NEIBOURS
multi
0,99
0,72
binary
0,99
0,77
4-class
0,99
0,74
GAUSSIAN
NAÏVE BAYES
multi
0,77
0,53
binary
0,84
0,55
4-class
0,65
0,42
MULTI LAYER
PERCEPTRON
multi
1,00
0,72
binary
1,00
0,79
4-class
1,00
0,77
25