Experimental Result 1
University of New South Wales,Sydney,Australia. 1999
Extracted rules:
Rule 1: (36/6, lift 1.6)
Min <= 0.00119555
-> class A [0.816]
Rule 2: (3, lift 1.6)
BowPos = a
Min > 0.00153141
-> class A [0.800]
Rule 3: (3, lift 1.6)
BowPos = b
Max > 0.00263044
-> class A [0.800]
Rule 4: (7/1, lift 1.6)
BowPos = c
Max > 0.0022552
Min <= 0.00149608
-> class A [0.778]
Rule 5: (2, lift 1.5)
Scale = down
BowPos = a
Max > 0.00210158
-> class A [0.750]
Rule 6: (10/2, lift 1.5)
BowPos = d
Ave <= 0.001729075
-> class A [0.750]
Rule 7: (12/1, lift 1.7)
BowPos = c
Min > 0.00149608
-> class B [0.857]
Rule 8: (11/1, lift 1.7)
BowPos = a
Max <= 0.00210158
Min > 0.00119555
Min <= 0.00153141
-> class B [0.846]
Rule 9: (10/1, lift 1.7)
BowPos = c
Max <= 0.0022552
Min > 0.00119555
-> class B [0.833]
Rule 10: (19/4, lift 1.5)
BowPos = b
Max <= 0.00263044
Min > 0.00119555
-> class B [0.762]
Rule 11: (8/2, lift 1.4)
Scale = up
BowPos = a
Min > 0.00119555
Min <= 0.00153141
-> class B [0.700]
Rule 12: (20/6, lift 1.4)
BowPos = d
Ave > 0.001729075
Min > 0.00119555
-> class B [0.682]
Default class: A
Evaluation on training data (128 cases):
Decision Tree Rules
---------------- ----------------
Size Errors No Errors
12 24(18.8%) 12 24(18.8%) <<
(a) (b) <-classified as
---- ----
49 15 (a): class A
9 55 (b): class B
Time: 0.1 secs
This result doesn’t make sense