Question

Based on the below data:

Customer1:Bread,Cereals,Milk

Customer2:Tomatoes,Eggs

Customer3:Pork,Bread,Milk

Customer4:Sugar,Tomatoes,Pork,Bread

Customer5:Vinegar

Customer6:Eggs,Milk,Cereals,Sugar,Pork

Customer7:Eggs,Milk,Vinegar

Customer8:Sugar,Pork

explain how the FP-tree can be used to mine frequent item sets and how it is different from the Apriori algorithm.

2.Apply the Apriori Method to the following dataset using excel using a support threshold of 20% and a lift threshold of 1.

3.Build an FP-tree using the following dataset( No need to generate the frequent itemsets).

Answer 1

1) In FP tree, each node denotes the item and the current count i.e., the number of times the item is used, and each branch of its denotes the association. Many bottlenecks of apriori were removed or addressed in the FP-tree.

2)

3) The FT tree for the data given would be as below, where the initial node would be null and the tree grows along with the transactions of different customers, the counts would be as given in 2)