Scribd
Upload
36 views

Decision Tree Program Results

A decision tree model was created using the HPSPLIT procedure to predict a target variable (TREG1) using a number of predictor variables. The model was built using a training dataset of 6,500 observations, with entropy as the split criterion and cost-complexity pruning. The optimal tree had 27 leaves. Performance statistics like the confusion matrix and AUC are reported. Variable importance measures the contribution of each predictor to the model.

Uploaded by

Chirag Saraogi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
36 views

Decision Tree Program Results

A decision tree model was created using the HPSPLIT procedure to predict a target variable (TREG1) using a number of predictor variables. The model was built using a training dataset of 6,500 observations, with entropy as the split criterion and cost-complexity pruning. The optimal tree had 27 leaves. Performance statistics like the confusion matrix and AUC are reported. Variable importance measures the contribution of each predictor to the model.

Uploaded by

Chirag Saraogi
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 6

S a t u r d a y , J u ly

T h e H P S P L IT P r o c e d u r e
P e r fo r m a n c e In fo r m a t io n
E x e c u t io n M o d e

S i n g le - M a c h i n e

N u m b e r o f T h r e a d s

D a t a A c c e s s In fo r m a t io n
D a ta

E n g in e

R o le

P a th

W O R K .N E W

V 9

In p u t

O n C li e n t

M o d e l In fo r m a t io n
S p lit C r it e r io n U s e d

E n tr o p y

P r u n in g M e t h o d

C o s t - C o m p le x i t y

S u b t r e e E v a lu a t io n C r it e r io n

C o s t - C o m p le x i t y

N u m b e r o f B r a n c h e s

M a x im u m

T r e e D e p th R e q u e s te d

1 0

M a x im u m

T r e e D e p t h A c h ie v e d

1 0

T r e e D e p th

N u m b e r o f L e a v e s B e fo r e P r u n in g

3 2 1

N u m b e r o f L e a v e s A ft e r P r u n in g

2 7

M o d e l E v e n t L e v e l

N u m b e r o f O b s e r v a t io n s R e a d

6 5 0 4

N u m b e r o f O b s e r v a t io n s U s e d

6 5 0 0

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

S a t u r d a y , J u ly

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

T h e H P S P L IT P r o c e d u r e

C o s t-C o m p le x ity A n a ly s is fo r T R E G 1 U s in g C ro s s V a lid a tio n


N u m b e r o f L e a v e s
1

1 4

1 8

3 5

4 4

5 1

7 9

9 8

1 0 9

1 5 3

1 7 8

1 7 9

1 8 5

2 3 2

2 4 3

A v e ra g e A S E

0 .1 5

0 .1 0

0 .0 5
1 S E
M in A v g A S E
N L e a v e s
P a ra m e te r

0 .0 0
96
.9

23

0.
0

08
8

0.
0

02
5

.0

0 .1 2 2
1 8
0 .0 0 0 7

.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
28
00
00
00
00
00
00
00
00
00
00
00
00
00
E 75
68
53
41
33
26
23
21
19
16
14
13
12
12

C o s t - C o m p le x it y P a r a m e t e r

S a t u r d a y , J u ly

T h e H P S P L IT P r o c e d u r e

C la s s ific a tio n T re e fo r T R E G 1

5
6

9
E

C
H

I
L

k
g

l
p

T R E G 1

q
2

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

S a t u r d a y , J u ly

T h e H P S P L IT P r o c e d u r e

S u b tre e S ta rtin g a t N o d e = 0

N o d e
N
2

m a re v e r1
1

m a re v e r1
0 o r M is s

N o d e
N
2

N o d e
N
2

W H IT E
1 o r M is s

W H IT E
0

N o d e
N
1

N o d e
N
2

c ig a v a il
1

c ig a v a il
0 o r M is s

N o d e
N
1

N o d e
N
2

A L C E V R 1
1
N o d e
N
2

D E V IA N T 1 D E V IA N T 1
< 4 .0 5 0 o r ... > = 4 .0 5 0
N o d e
N
2

T R E G 1 = 1

N o d e
N
2

A L C E V ...
0 o r M is s
N o d e
N
2

W H IT E
1 o r M is s

W H IT E
0

N o d e
N
2

N o d e
N
2

T R E G 1 = 2

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

S a t u r d a y , J u ly

T h e H P S P L IT P r o c e d u r e
M o d e l- B a s e d C o n fu s io n M a t r ix
P r e d ic t e d
1

E r r o r
R a te

6 3 2

6 5 3

0 .5 0 8 2

2 9 8

4 9 1 7

0 .0 5 7 1

A c tu a l

M o d e l- B a s e d F it S t a t is t ic s fo r S e le c t e d T r e e
N
L e a v e s

A S E

M is c la s s

E n tr o p y

G in i

R S S

2 7

0 .1 0 9 6

0 .1 4 6 3

0 .5 1 3 8

0 .2 1 9 2

1 4 2 4 .8

R O C C u rv e fo r T R E G 1
1 .0

0 .8

S e n s itiv ity

0 .6

0 .4

0 .2

T r a in in g A U C

0 .0
0 .0

0 .2

0 .4

0 .6

1 - S p e c if ic ity
T r a in in g

0 .8

0 .8 4
1 .0

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

S a t u r d a y , J u ly

T h e H P S P L IT P r o c e d u r e
V a r ia b le Im p o r t a n c e
T r a in in g
V a r ia b le

V a r ia b le
L a b e l

R e la t iv e

Im p o r t a n c e

C o u n t

1 .0 0 0 0

1 8 .8 7 7 0

0 .4 7 3 3

8 .9 3 5 3

A L C E V R 1

0 .3 4 5 2

6 .5 1 6 2

D E V IA N T 1

0 .3 1 7 5

5 .9 9 2 9

c ig a v a il

0 .2 5 5 1

4 .8 1 5 6

G P A 1

0 .2 2 2 1

4 .1 9 2 4

A L C P R O B S 1

0 .2 1 8 4

4 .1 2 3 3

S C H C O N N 1

0 .1 9 2 6

3 .6 3 4 8

0 .1 7 4 6

3 .2 9 6 4

D E P 1

0 .1 5 2 4

2 .8 7 7 7

P A R P R E S

0 .1 2 7 6

2 .4 0 7 9

F A M C O N C T

0 .1 2 6 6

2 .3 8 9 8

P A S S IS T

0 .1 1 9 8

2 .2 6 1 5

c o c e v e r 1

0 .1 1 7 6

2 .2 1 9 6

a g e

0 .1 0 5 8

1 .9 9 6 6

0 .0 9 8 3

1 .8 5 5 0

0 .0 9 8 1

1 .8 5 2 2

m a r e v e r 1
W H IT E

H IS P A N IC

A S IA N
P A R A C T V

S 1 Q 6 A R A C E -W H IT E -W 1

S 1 Q 4 A R E Y O U O F H IS P A N IC O R IG IN -W 1

S 1 Q 6 D R A C E -A S IA N -W 1

2 , 2 0 1 6 0 7 :5 9 :3 2 P M

You might also like