Association rule mining

Question 1

Associative rule mining

Answer 1

is an unsupervised learning technique used to discover interesting relationships between items in large datasets, particularly in market basket analysis.

Question 2

Mining Frequent Itemsets

Answer 2

Finding combinations of items that occur together frequently in transactions

Question 3

The Apriori Algorithm

Answer 3

Principle: If an itemset is frequent, all its subsets must also be frequent (anti-monotone property)

• The support of an itemset never exceeds the support of its subsets
• This is known as the anti-monotone property of support

Question 4

Apriori algorithm steps

Answer 4

How it works (Step-by-step):

Find frequent individual items

Count how often each item appears.

Keep only those that meet the minimum support threshold.

Generate larger itemsets

Combine frequent items to form 2-item sets, 3-item sets, and so on.

Keep only those that are still frequent (meet support threshold).

Stop when no larger frequent sets can be made.

Generate rules from frequent itemsets

For each frequent itemset, create rules like A → B.

Keep only the rules with high confidence (above the threshold).

Question 5

Confidence

Answer 5

Confidence measures how often the rule is true. eg Given that someone buys beer, how often do they also buy wine?

Question 6

Drawback of confidence

Answer 6

Confidence can be misleading because it doesn’t consider how common the right-hand side item already is. For example, even if P(Coffee | Tea) = 0.75, but P(Coffee) = 0.9, buying tea actually reduces the likelihood of buying coffee.

Question 7

what are closed and maximal sets?

from these transactions which are closed and maximal

Answer 7

An itemset is closed if no proper superset has the same support.
An itemset is maximal if it is frequent and has no frequent superset.

Question 8

frequent?

Answer 8

Frequent = support ≥ min_sup