association-rule

COMP 4331 Knowledge Discovery and Data

Mining, Fall 2018

Frequent Itemset Mining via Programming

There is a transacton database stored in the link: https://github.com/comp4331fall18/sample-code-DM/blob/master/freq_items_datasets.txt.

Mine Frequent Itemsets

• Frequent itemset mining using Apriori of the transactions (Compile Apriori method in the A1 itsc stuid code Apr.py)

• Frequent itemset mining based on FP-Growth (Compile FP-growth methodin the A1 itsc stuid code fp.py)

• Frequent itemset mining via Recursive Elemination (Compile Relim methodin the A1 itsc stuid code relim.py).

Closed Frequent Itemsets

Based on the frequent itemsets you mined previously, write a Python 2 or Python 3 program to mine the closed frequent itemsets and maximal frequent itemsets.

Data Description

The benchmark dataset freq items dataset is supported by the IBM Almaden Quest research group, which contains 1,000 items and 100,000 transactions. For simplicity, each number uniquely identifies an item.