Advanced AI and ML - CIE 3 sol

Department of Computer Science and Engineering
(Artificial Intelligence and Machine Learning)
Continuous Internal Evaluation Test III - ODD Semester 2025 - 26
Course Title: Advanced AI and ML Course Code: AM722I1A

Solution
Module 4
Q. No. Solution Marks
a.




6




1



b.




5

, c. Three core metrics:

Support(A → B) = fraction (or count) of transactions containing A ∪ B. Measures
how frequently the rule items occur together.

Confidence(A → B) = support(A ∪ B) / support(A). Estimates probability that B
appears when A appears. (Conditional probability P(B|A)).

Lift(A → B) = confidence(A → B) / support(B). Measures how much more likely B is
to occur with A than by random chance; lift > 1 indicates positive association.

Market-basket use & tiny recommendation example:
Suppose transactions show {Bread, Butter} appears in 40% of baskets (support=0.40). Bread
appears in 50% of baskets (support(Bread)=0.50). Butter appears in 45%
(support(Butter)=0.45). 4

support(Bread→Butter) = 0.40

confidence(Bread→Butter) = 0..50 = 0.80 → when a customer buys Bread, 80%
also buy Butter.

lift(Bread→Butter) = 0..45 ≈ 1.78 → customers buying Bread are 1.78× more
likely to buy Butter than average.

Recommendation: If a shopper has Bread in cart, recommend Butter (high confidence and
lift). In a rudimentary recommender, present rules with high confidence & lift; can also rank
suggestions by lift or confidence.