You are on page 1of 5

EX.

NO:5 FINDING ASSOCIATION RULES USING WEKA TOOL


21-Mar-2017.
Loading vote dataset

Applying apriori algorithm to find the support and confidence limit of various association
rules.
Identifying positive instances for different support counts
low support count:0.1
high support count:1

Low support count:0.5


high support count:1.5
Support = Count (L) / No. of transactions Rule
Confidence = Support Count (L, R)/Support Count(L)

Lift (XY) = Support (X U Y) / (Support (X) x Support (Y))

Conviction (XY) = (Sup (X) x Sup (~Y)) / Sup (X, ~Y)

Leverage (XY) = Sup (X, Y) (Sup (X) x Sup (Y))

E.g.: outlook=overcast 4 ==> play=yes 4 lift:(1.56) lev:(0.1) [1] conv:(1.43)


Confidence =4/4 = 1 Lift = (4/14) / ((4/14) x (9/14)) = 1.555 Leverage = (4/14)
((4/14) x (9/14)) = 0.1021

No of rules changed to be 4

Applying apriori algorithm to the dataset WEATHER.NOMINAL and checking the support and
confidence of a particular rule.
Making association rules for SUPERMARKET dataset with the number of rules as 30.
RESULT:-
Thus association rules for a particular dataset has been computed successfully using
weka tool.

You might also like