You are on page 1of 6

9/1/2020 [2ICT703] Machine Learning

[2ICT703] Machine Learning


The name and photo associated with your Google account will be recorded when you upload
files and submit this form. Not abhishekdixit17@gnu.ac.in? Switch account

Any files that are uploaded will be shared outside of the organization they belong to.

* Required

MCQ

All question are Mandatory


No negative marks
15 MCQ Questions (Some of them might be having more than one option)

Q1. A and B are two events. If P(A, B) decreases while P(A) increases, which 1 point

of the following is true? *

A) P(A|B) decreases

B) P(B|A) decreases

C) P(B) decreases

D) All of above

Q2. Choose the correct option(s) from the following. * 1 point

A) When working with a small data set, one should prefer low bias/high variance
classifiers over high bias/low variance classifiers.

B) When working with a small data set, one should prefer high bias/low variance
classifiers over low bias/high variance classifiers.

C) When working with a large data set, one should prefer high bias/low variance
classifiers over low bias/high variance classifiers.

D) When working with a large data set, one should prefer low bias/high variance
classifiers over high bias/low variance classifiers.

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 1/6
9/1/2020 [2ICT703] Machine Learning

Q3. After applying a regularization penalty in linear regression, you find 1 point

that some of the coefficients of w are zeroed out. Which of the following
penalties might have been used? *

a) L1 norm

b) L2 norm

c) either (a) or (b)

d) Neither of them can do this

Q4. Suppose we like to calculate P(H|E) and we have no conditional 1 point

independence information. Which of the following sets of numbers are


sufficient for the calculation? *

a) P(E), P(H), P(E|H), P(H)

b) P(E), P(H), P(E, H)

c) P(H), P(E|H), P(H)

d) P(E), P(E|H), P(H)

Q5. Averaging the output of multiple decision trees helps ________. * 1 point

a) Increase bias

b) Decrease bias

c) Increase variance

d) Decrease variance

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 2/6
9/1/2020 [2ICT703] Machine Learning

Q6. Which of the following is more appropriate to do feature selection? * 1 point

a) Ridge

b) Lasso

c) both (a) and (b)

d) neither (a) nor (b)

Q7. Given data x which consists of x1,x2,.....xn vectors of d dimension, the 1 point

cost function to minimize squared error is given as, u ∈ Rd | x1 - u |^2 + | x2 -


u |^2 + |x3 - u |^2+......|xn - u |^2 what is the closed-form expression for u? *

A) mean(x)

B) median(x)

C) variance(x)

D) None of the above

Q8. Let v be the eigenvector and eigenvalue λ for equation Av = λv, where 1 point

A is a square matrix. If given matrix is now scaled up to A50, does v still


become an eigenvector for that scaled matrix? if yes, what would be
eigenvalue for it? *

A) Yes, λ + 50

B) Yes,λ50

C) Yes,50λ

D) No,Cannot determine

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 3/6
9/1/2020 [2ICT703] Machine Learning

Q9. Consider three-layer fully connected neural-network architecture 1 point

having (64,200,10) neurons in the first, second, and third layers. Calculate
the total number of multiplications, additions happened for learnable
parameters (w,b). Note: calculations for activation functions are not to be
counted. *

A) 14800, 14798

B) 128000,127998

C) 13000,12998

D) 12800,12798

Q10. Answer in next cell

Q10. * 1 point

||x||2 ≤ ||x||1 is true and ||x||1 ≤ ||x||2 is false

||x||2 ≤ ||x||1 is false and ||x||1 ≤ ||x||2 is true

||x||2 ≤ ||x||1 and ||x||1 ≤ ||x||2 are both false

insufficient information

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 4/6
9/1/2020 [2ICT703] Machine Learning

Q11. Given data X: {x1,x2,x3….xn} of 4 dimension each, we want to reduce to 1 point

2 dimension. Given eigen values are λ1 = 6, λ2 = 2, λ3 = 0, λ4 = 6, which


eigen vector are taken as PCA components? Note: v1,v2,v3,v4 are eigen
vectors with eigen values (λ1, λ2,λ4,λ3) respectively *

A) v1,v4

B) v2,v3

C) v1,v3

D) v3,v4

Q12. Suppose you are playing a game Get Lucky Lottery where a randomly 1 point

generated number at every roll from 1-100 inclusive is given out, if that
number is above 35 you win some amount x, else you need to pay the
same amount x. Which option you would chose to maximize your expected
gain? *

A. Play 100 times, betting $1 each time.

B. Play 10 times, betting $10 each time.

C. Play one time, betting $100.

D. All of the above

Q13. 3 friends each choose a random number between 1 and 4. What is the 1 point

chance that any of them chose the same number? *

A) 1/16

B) 1/4

C) 8/16

D) 7/16

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 5/6
9/1/2020 [2ICT703] Machine Learning

Q14. Which loss function is widely used in Linear Regression? * 1 point

A) LogLoss

B) Maximum Likelihood Loss

C) Mean-Squared error

D) Mean Bias Error

Q15. Which of the following is not true about Random Forest training? (This 1 point

may have more than one answer, select all that apply) *

A) samples which are not used in construction of individual trees are made to run
through the trees to check the training accuracy using major votes prediction

B) While constructing a tree, same samples can be repeated multiple times in


bootstrap data

C) Number of features require to split the node is selected at random

D) None of the these

Page 3 of 4

Back Next

Never submit passwords through Google Forms.

This form was created inside of Ganpat University. Report Abuse

 Forms

https://docs.google.com/forms/d/e/1FAIpQLSeuUdCJsHcuojUT88_7GvJ1O_fvmbHlZDJWKqOuS-KsJOlzgQ/formResponse 6/6

You might also like