MODEL QUESTION PAPER: 2024-25
Course Code: BCSML601
VIDYAVARDHAKA COLLEGE OF ENGINEERING
Autonomous Institute, Affiliated to Visvesvaraya Technological University, Belagavi
Gokulam, 3rd Stage, Mysuru 570 002
Sixth Semester BE Examinations
COURSE NAME: MACHINE LEARNING
Duration: 3 hours Max. Marks: 100
INSTRUCTION TO STUDENTS
1) Answer One Full question from each module.
2) Three module questions are compulsory and the remaining two modules will have internal choice.
3) Each module carries 20 marks.
Q.
No. Module-I Marks BL CO/
PO
1. (a) Discuss different types of machine learning techniques with an example. 10 L2 CO1/
PO1
1. (b) Analyze the following scenarios w.r.t implementation of machine learning 10 L4 CO3/
algorithms and answer the following questions- PO2
i) Recommend suitable machine learning techniques (supervised,
unsupervised and reinforcement learning algorithm)
ii) Justify and state the process briefly on how to implement the machine
learning technique for the task
Scenario-1
Students attend classes in which cameras are installed, students’
attendance for each class is recorded through the camera by capturing their
faces. The faculty in charge for the class uses an automated interface to
capture image of the class to report the attendance.
Scenario-2
Assume you want to purchase an exquisite gift item for your friend’s
birthday. The gift item is only available in one of the stores in the city. You
don’t know where the stores are located in the city. The birthday is 4 days
ahead from now. Customers who visit stores are entertained by the store
owners.
Module-II
2.(a) Explain the process of multi-class classification with a suitable example. 10 L2 CO1/
PO1
2.(b) Illustrate the process of Optimal Separating Hyperplane for a binary 10 L3 CO2/
classification problem. PO1
Module-III
1
MODEL QUESTION PAPER: 2024-25
Course Code: BCSML601
3.(a) Discuss the appropriate problems for learning neural network. 10 L2 CO1/
PO1
3.(b) Illustrate back propagation algorithm using stochastic gradient descent 10 L3 CO2/
method. PO1
OR
4.(a) Explain the appropriate problems for neural network learning. 10 L2 CO1/
PO1
4.(b) Design a two-input perceptron that implements the following Boolean 10 L3 CO2/
function (A ^ B) by considering w1=1.2, w2=0.6, threshold=1 and learning PO1
rate = 0.5.
Module-IV
Consider the car theft problem with attributes Color, Type, Origin, and the
10 L3 CO2/
5.(a) target, Stolen can be either Yes or No. Estimate the conditional probabilities
PO1
values for the instance {Color=Red, Type=SUV, Origin = Domestic}. Refer
Table 5a.
Table 5a. Car Theft
5.(b) Analyze Minimum Description Length Principle for choosing hypothesis to 10 L4 CO3/
minimize sum of two description lengths. PO2
OR
6.(a) 10 L3 CO2/
Illustrate Bayes Theorem and calculate the probability of a dangerous fire
PO1
when there is Smoke, Given:
a) Dangerous fires are rate (1%)
b) Smoke is fairly common (10%) due to barbecues
c) 90% of dangerous fires make smoke
6.(b) Prove, how maximum likelihood (Bayesian Learning) can be used in any 10 L4 CO3/
learning algorithms that are used to minimize the squared error between PO2
actual output hypothesis and predicted output hypothesis.
Module-V
7.(a) Explain the learning mechanism in Reinforcement Learning with a neat 10 L2 CO1/
diagram. PO1
7.(b) Illustrate the usage of K Nearest Neighbor classifier to predict the diabetic 10 L3 CO2/
2
MODEL QUESTION PAPER: 2024-25
Course Code: BCSML601
patient with the given features BMI and Age. PO1
Assume K=3, for the given test example BMI=43.6, Age=40, Sugar =? Refer
to Table 7b.
Table 7b: Diabetic patient
BMI Age Sugar
33.6 50 1
26.6 30 0
23.4 40 0
43.1 67 0
35.3 23 1
35.9 67 1
36.7 45 1
25.7 46 0
23.3 29 0
31 56 1
