Question: Use the Apriori to algorithm to identify the frequent item-sets in the following database. Then extract the strong association rules from these sets.
0

Min. Support = 30% Min. Confidence=75%

TID Items
01 A, B, D, E, F
02 B, C, E
03 A, B, D, E
04 A, B, D, E
05 A, B, C, D, E, F
06 B, C, D
07 A, B, D, E
dmbi(26) • 358 views
ADD COMMENTlink
modified 8 months ago  • written 9 months ago by gravatar for ASHISH RAVINDRA SALVE ASHISH RAVINDRA SALVE10
0

Given:-

TID Items
01 A, B, D, E, F
02 B, C, E
03 A, B, D, E
04 A, B, D, E
05 A, B, C, D, E, F
06 B, C, D
07 A, B, D, E

Solution:-

Step 1: Generating Item set(support count = no of occurrences of item)

Item Support
A 5
B 7
C 4
D 5
E 6
F 2

filtering item set with 30% minimum support(i.e. 30% of total transaction =2.3)

Item Support
A 5
B 7
C 4
D 5
E 6
F 2

Step 2: Generating 2 Item set

Item Support
AB 5
AC 2
AD 4
AE 5
BC 4
BD 4
BE 6
CD 2
CE 3
DE 4

filtering item set with 30% minimum support

Item Support
AB 5
AD 4
AE 5
BC 4
BD 4
BE 6
CE 3
DE 4

Step 43: Generating 3 Item set

Item Support
ABC 2
ABD 4
ABE 5
ADE 4
BCD 2
BCE 2
BDE 4

filtering item set with 30% minimum support

Item Support
ABD 4
ABE 5
ADE 4
BDE 4

Step 4:- Generating 4 Item set

Item Support
ABDE 4
BDEF 2
CDEF 1

filtering item set with 30% minimum support

Item Support
ABDE 4

Step 5; Generating rules and confidence

Rule Confidence Confidence percentage
A->BED 4/5=0.8 80%
B ->AED 4/7=0.57 57%
E -> ABD 4/5=0.8 80%
D ->ABE 4/6=0.66 66%
AB->ED 4/5=0.8 80%
BE->AD 4/6=0.66 66%
ED ->AB 4/4=1 100%
AE->BD 4/5=0.8 80%
AD->BE 4/4=1 100%
BED->A 4/4=1 100%
AED ->B 4/4=1 100%
ABD->E 4/4=1 100%
ABE->D 4/5=0.8 80%

From the above Rules generated, only the rules having greater than 75% confidence are considered as final rules. So Final rules are

A->BED

E -> ABD

AB->ED

ED ->AB

AE->BD

AD->BE

BED->A

AED ->B

ABD->E

ABE->D

ADD COMMENTlink
written 8 months ago by gravatar for ASHISH RAVINDRA SALVE ASHISH RAVINDRA SALVE10
Please log in to add an answer.