DSCI EXAM 4520 EXAM 3 QUESTIONS WITH VERIFIED
ANSWERS
Assume that for a user-based collaborative recommender system the similarity scores
are being calculated for five customers who have rated five brands of tea products:
Ratings of which two customers are better to be used for predicting Customer 5's rating
of the Celestial Tea products?
- Customers 2 and 3
- Customer 4 and 1
- Customer 3 and 1
- Customer 2 and 1 - Answers - Customers 2 and 3
Which of the followings is NOT an application of the market basket analysis?
-Predicting customers spending amount
-Bundling of products and services
-Product placement in retail store establishments
- Recommendation to customers - Answers - Predicting customers spending amount
In association rule extraction, lift measures what percentage of baskets that contained
the left-hand product also contained the right-hand products. T/F - Answers - False
What similarity function we used in this course to quantify user similarities in the User-
Based Collaborative Filtering model?
-Pearson Correlation Coefficient Function
-Euclidean Distance Function
-Manhattan Distance Function
-Absolute Similarity Function - Answers - Pearson Correlation Coefficient Function
In market basket analysis, support refers to the percentage of baskets where the rule is
true (i.e. both left and right-side products are in the basket). T/F - Answers - True
Which statement is correct about support, confidence, and lift in basket market
analysis?
- Rules with higher confidence values are more effective and desirable
-Support can be between -1 and 1
-Rules with lower lift values are more effective and desirable
ANSWERS
Assume that for a user-based collaborative recommender system the similarity scores
are being calculated for five customers who have rated five brands of tea products:
Ratings of which two customers are better to be used for predicting Customer 5's rating
of the Celestial Tea products?
- Customers 2 and 3
- Customer 4 and 1
- Customer 3 and 1
- Customer 2 and 1 - Answers - Customers 2 and 3
Which of the followings is NOT an application of the market basket analysis?
-Predicting customers spending amount
-Bundling of products and services
-Product placement in retail store establishments
- Recommendation to customers - Answers - Predicting customers spending amount
In association rule extraction, lift measures what percentage of baskets that contained
the left-hand product also contained the right-hand products. T/F - Answers - False
What similarity function we used in this course to quantify user similarities in the User-
Based Collaborative Filtering model?
-Pearson Correlation Coefficient Function
-Euclidean Distance Function
-Manhattan Distance Function
-Absolute Similarity Function - Answers - Pearson Correlation Coefficient Function
In market basket analysis, support refers to the percentage of baskets where the rule is
true (i.e. both left and right-side products are in the basket). T/F - Answers - True
Which statement is correct about support, confidence, and lift in basket market
analysis?
- Rules with higher confidence values are more effective and desirable
-Support can be between -1 and 1
-Rules with lower lift values are more effective and desirable
