21AI71

Model Question Paper for 2021(CBCS Scheme)

USN

Seventh Semester B.E. Degree Examination

21AI71
ADVANCED AI&ML
TIME: 03 Hours Max. Marks: 100

Note:Answer any FIVE full questions, choosing at least ONE question from each MODULE.

*Bloom’s
Module -1 Taxonomy COs Marks
Level
Q.01 a L1 CO1 10
What is Artificial Intelligence ?, Discuss the branches of AI.
b L2 CO1 10
What is state space ? Explain the concept of state space representation
using the water Jug Problems.
OR
Q.02 L2 CO2 10
a Explain any two AI techniques for solving tie-tar-toe problems.
L3 CO1 10
b Write the algorithm for breadth first search and depth-first . Enlist the
advantage of each.
Module-2
Q. 03 a Explain the properties of good knowledge representation systems. L2 CO2 10

L4 CO2 10
b Define the following terms 1. Concept learning ii. Consistent hypothesis
iii. Version space
OR
Q.04 L2 CO2 10
a Define uncertainty, explain the effects of practical ignorance.
L1 CO2 10
b Explain inference using full joint distribution
Module-3
Q. 05 L4 CO3 10
a Design the perceptron that implement AND functions. Why is that a single
layer perceptron can not be used to represent XOR function.
L2 CO3 10
b Derive an equation for gradient discrete rule to minimize the error.
OR
Q. 06 a Write a short note on Scrum and Crystal. L3 CO3 10

b Explain the core principles and practices of software engineering L2 CO3 10

Module-4
Q. 07 L4 CO4 10
a Define maximum likelihood hypothesis derive an equation for ML
hypothesis using Bayes theorem
L1 CO4 10
b Given a user-item interaction matrix with 5 users and 4 items, how many
latent factors would you need if you're using Singular Value
Decomposition (SVD) for matrix factorization? Assume you choose 3
latent factors.
OR
Page 01 of 02
 21AI71
Q. 08 a If a user has interacted with 3 items with ratings 4, 3, and 5 respectively, L2 CO4 10
and the model predicts ratings of 4.2, 2.9, and 4.8, calculate the Mean
Squared Error (MSE) between the predicted and actual ratings.
b Describe a hybrid recommender system that combines collaborative L3 CO4 10
filtering and content-based filtering. How can these two approaches
be combined, and what are the potential challenges, such as data
sparsity and scalability?

Q-09 a. You have a user-item rating matrix where User A has rated Items 1, 2, and L5 CO5 10
3 with values of 5, 3, and 4, respectively. User B has rated Items 1, 3, and 4
with values of 2, 5, and 4, respectively. Using cosine similarity, calculate
the similarity score between User A and User B.
b. consider a small dataset with 6 data points and two clusters: L6 CO6 10

Calculate the silhouette score for one point from each cluster (say P1 from
Cluster 1 and P4 from Cluster 2).
OR
Q-10 a. How does the choice of distance metric impact the performance of L5 CO5 10
clustering algorithms in unsupervised learning?
b. What is the "curse of dimensionality," and how does it affect clustering L6 CO6 10
algorithms, particularly in high-dimensional spaces?

Page 01 of 02
 21AI71

Module-5
Q. 09 a Define software quality and explain place of software quality in project L2 CO5 10
management.

L2 CO5 10
b Explain Capability Maturity Model and CMM key areas.
OR
Q. 10 a Explain Product v/s Process quality management. L2 CO5 10

L1 CO5 10
b Explain in detail about Empirical Estimation Model
*Bloom’s Taxonomy Level: Indicate as L1, L2, L3, L4, etc. It is also desirable to indicate
the COs and POs to beattained by every bit of questions.

Page 01 of 02