MTS 488 - Lesson 5

What is the other name of "Association Rule Mining"?

• Matching Strategy

• Market Basket Analysis

• The Marketing Mix

• Basket Market Analysis

What is NOT true about "Association Rule"?

• It's supervised learning

• X → Y is an implication form of it

• It's unsupervised learning

• It does not mean buying X is the reason to buy Y.

What is the meaning of X → Y?

• Implication of causality

• X is Y's consequent.

• If customers buy X, they will likely to buy Y.

• More than 1 answers are corrected.

What is NOT the main metrics of an association rule?

• Lift

• Interval ratio

• Confidence

• Support

What is NOT correct about the support?

• The high support indicates the rule is frequent enough to be interested.

• Support quantifies the significance of the co-occurance of a rule.

• Number of transactions involving a particular itemset.

• It depends on the frequency of ‘Y’ appearing solely.

What can we tell about the rule from having support = (4/12) and confidence = 66%?

• No association

• Negative association

• Positive association

• The data is not enough.

What is the frequency of (Cornflakes, Jam)?

• 2

• 4

• 3

• 1

What is the confidence value of Water→Bread?

• 100%

• 67%

• 33%

• 50%

Which is NOT the way to store datasets of transactions?

• Simple Storage

• Horizontal Storage

• Vertical Storage

• Semi-simple Storage

What is true about Apriori Algorithm?

• Best to identify the frequent items in the dataset and ignore all others.

• Consist of 3 phases

• It generates candidate itemsets and then tests them.

• Have a limitation with large datasets.