Question 1

A grocery store introducing items from Saudi Arabia is interested in analyzing buying trends of these new "international" items, namely Y, M, C, and J. The files Grocery Store List and Grocery Store Stacked provide data on a collection of transactions in item-list format. Biscuit = B, Soda = S, Almari Cheese = C, Avocado = A, Olives = O, Black Olives = D, Almari Yogurt = Y, Crackers = E, Almari Juices = J, Almari Milk = M, Mutton = F, Coke = L, Apples = H, Chicken = I, Peppers = P, Beef = K, Turkey = T, Gelato = G, Steak = N.

serial number Grocery Store list______________________________________

1 B Y O K J T G

2 C S B E M O Y

3 A E F K M T D

4 O J L T G K P

5 B Y H O N A T

6 D M I L G P K

7 O J L T M G H

8 Y J P E I G C

9 S J O E M P C

10 Y J P E I K B

11 C D H P A N T

12 C B A F M H Y

a) Use a minimum support of 4 transactions (33.33% of the 12 total transactions) and a minimum confidence of 50% to generate a list of association rules. How many rules satisfy this criterion?

b) Use a minimum support of 8 transactions (66.66% of the 12 total transactions) and a minimum confidence of 50% to generate a list of association rules. How many rules satisfy this criterion?

c) Why may the grocery store want to increase the minimum support required for their analysis? What is the risk of increasing the minimum support required?

d) Using the list of rules from part (b), consider the rule with the largest lift ratio that also involves a Saudi item. Interpret what this rule is saying about the relationship between the antecedent item set and consequent item set.