ITEC 451 Mining Frequent Patterns, Association, and Correlations

Frequent Pattern Analysis

- A pattern (a set of items, subsequences, substructures, etc.) that occurs frequently in a data set. - Applications: Basket data analysis, cross-marketing, catalog design, sale campaign analysis, Web log (click stream) analysis, and DNA sequence analysis.

What is a closed itemset and what is the downward closure property?

- An itemset is closed if none of its immediate supersets has the same support as the itemset. - The downward closure property of frequent patterns: Any subset of a frequent itemset must be frequent. If {beer, diaper, nuts} is frequent, so is {beer, diaper} and {beer,nuts}.

What is an association rule?

- Find all the rules X  Y with minimum support and confidence - Support, s, probability that a transaction contains X union (u) Y - Confidence, c, conditional probability that a transaction having X also contains Y.

What is a frequent itemset in a dataset?

- Frequently Bought Together - Customers who bought this item also bought...

What is the Apriori pruning principle and how does it make the algorithm more efficient?

- If there is any itemset which is infrequent, its superset should not be generated/tested! - Initially, scan DB once to get frequent 1-itemset. - Generate length (k+1) candidate itemsets from length k frequent itemsets. - Test the candidates against DB. - Terminate when no frequent or candidate set can be generated.

Properties of frequent itemsets.

- k-itemset X = {x1, ..., xk} - (Absolute) support, or, support count of X: Frequency or occurrence of an itemset X - (Relative) support, s, is the fraction of transactions that contains X (i.e., the probability that a transaction contains X).

the Apriori algorithm work? Be able to demonstrate on a set of frequent itemsets.

1. Count the number of transactions in which each item occurs. 2. Let's say that an item is frequent if it occurs in 60% of the transactions. In step 2 we simply remove any items that are bought less than 3 times. 3. Start making pairs from the first item (M) and the second item (O) and so on... 4. Count how many times each pair appears together in a transaction 5. Remove all the L2 transaction pairs that occur less than 3 times and we are left with the following. 6. Form sets of three items using the self join rule. For each item pair we find two items with the same first item and join them (OK and OE = OKE).