5/12/2020 MST-2_Machine Learning

MST-2_Machine Learning
* Required

Untitled Section

18. Support vector machine (SVM) is a _________ classifier? * 1 point

Discriminative

Generative

41. How many input nodes are required to process a grayscale image of 1 point
28X28? *

28 X 1

28 X 28

56 X 56

56 X 1

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 1/14
5/12/2020 MST-2_Machine Learning

15. What is the objective of backpropagation algorithm? * 1 point

to develop learning algorithm for multilayer feedforward neural network

to develop learning algorithm for single layer feedforward neural network

to develop learning algorithm for multilayer feedforward neural network, so that

none of the mentioned

6. What is reinforcement learning? * 1 point

learning is based on evaluative signal

learning is based o desired output for an input

learning is based on both desired output & evaluative signal

none of the mentioned

34. When deriving a memory cell in memory networks, we choose to read 1 point
values as vector values instead of scalars. Which type of addressing would
this entail? *

Content-based addressing

Location-based addressing

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 2/14
5/12/2020 MST-2_Machine Learning

4. What is purpose of Axon? * 1 point

receptors

transmitter

transmission

none of the mentioned

33. A recurrent neural network can be unfolded into a full-connected 1 point

TRUE

FALSE

24. The goal of the SVM is to __________ * 1 point

Find the optimal separating hyperplane which minimizes the margin of training data

Find the optimal separating hyperplane which maximizes the margin of training data

11. In nearest neighbour case, the stored pattern closest to input pattern is 1 point
recalled, where does it occurs? *

feedback pattern classification

feedforward pattern classification

can be feedback or feedforward

none of the mentioned

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 3/14
5/12/2020 MST-2_Machine Learning

21. SVM is termed as ________ classifier * 1 point

Minimum margin

Maximum margin

39. In neural networks, nonlinear activation functions such as sigmoid, tanh, 1 point

and ReLU *

speed up the gradient calculation in backpropagation, as compared to linear units are

help to learn nonlinear decision boundaries always output values between 0 and 1

are applied only to the output units

always output values between 0 and 1

3. Function of dendrites is? * 1 point

receptors

transmitter

both receptor & transmitter

none of the mentioned

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 4/14
5/12/2020 MST-2_Machine Learning

28. The following is not an assumption of Markov analysis. * 1 point

There is an infinite number of possible states.

The probability of changing states remains the same over time.

We can predict any future state from the previous state and the matrix of transition
probabilities.

The size and composition of the system do not increase.

A and D

40. Which of the following types of RNN is best suited for sentiment 1 point

analysis? *

One to One

b) One to Many

c) Many to One

d)Many To Many

29. The network that involves backward links from output to the input and 1 point

hidden layers is called *

Self organizing maps

Perceptrons

Recurrent neural network

Multi layered perceptron

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 5/14
5/12/2020 MST-2_Machine Learning

25. Which of the following SVM model can be used to detect the outliers? * 1 point

nu-SVM classification

nu-SVM regression

one-classification

None of the above

20. SVM is a ___________ learning algorithm * 1 point

Supervised

Unsupervised

42. Which of the following is FALSE about Dying ReLU? * 1 point

In this situation, neurons become inactive and only output 0 for any input

Dying ReLU neurons do not play any role in discriminating the input and become
useless

Dying ReLU problem occurs when learning rate is too low

None of the above

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 6/14
5/12/2020 MST-2_Machine Learning

32. Exploding gradient problem is an issue in training deep networks where 1 point
the gradient getS so large that the loss goes to an infinitely high value and
then explodes.What is the probable approach when dealing with
“Exploding Gradient” problem in RNNs? *

Use modified architectures like LSTM and GRUs

Gradient clipping

Dropout

None of these

31. Given an n-character word, we want to predict which character would 1 point
be the n+1th character in the sequence. For example, our input is
“predictio” (which is a 9 character word) and we have to predict what
would be the 10th character.Which neural network architecture would be
suitable to complete this task? *

Fully-Connected Neural Network

b) Convolutional Neural Network

c) Recurrent Neural Network

d) Restricted Boltzmann Machine

19. SVM can be used to solve ___________ problems. * 1 point

Classification

Regression

Clustering

Both Classification and Regression

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 7/14
5/12/2020 MST-2_Machine Learning

16. What is true regarding backpropagation rule? * 1 point

it is a feedback neural network

actual output is determined by computing the outputs of units for each hidden layer

hidden layers output is not all important, they are only meant for supporting input and
output layers

none of the mentioned

7.How many types of reinforcement learning exist? * 1 point

14. What is the objective of perceptron learning? * 1 point

class identification

weight adjustment

adjust weight along with class identification

none of the mentioned

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 8/14
5/12/2020 MST-2_Machine Learning

8. What leads to minimization of error between the desired & actual 1 point
outputs? *

stability

convergence

either stability or convergence

none of the mentioned

23. Which of the following is a type of SVM? * 1 point

Maximum margin classifier

Soft margin classifier

Support vector regression

All of the above

26. Markov analysis is a technique that deals with the probabilities of future 1 point
occurrences by *

using Bayes' theorem.

analyzing presently known probabilities.

time series forecasting.

the maximal flow technique.

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 9/14
5/12/2020 MST-2_Machine Learning

27. Markov analysis might be effectively used for * 1 point

technology transfer studies.

university retention analysis.

machine breakdowns.

accounts receivable analysis.

all of the above

13. By using only linear processing units in output layer, can a artificial 1 point
neural network capture association if input patterns is greater then
dimensionality of input vectors? *

YES

NO

12. Feedforward networks are used for? * 1 point

pattern mapping

pattern association

pattern classification

all of the mentioned

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 10/14
5/12/2020 MST-2_Machine Learning

35. If calculation of reset gate in GRU unit is close to 0, which of the 1 point
following would occur? *

Previous hidden state would be ignored

Previous hidden state would be not be ignored

2. What is shape of dendrites like * 1 point

OVAL

ROUND

TREE

RECTANGULAR

9.What happens during recall in neural networks? * 1 point

weight changes are suppressed

input to the network determines the output activation

both process has to happen

none of the mentioned

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 11/14
5/12/2020 MST-2_Machine Learning

30. Why is an RNN (Recurrent Neural Network) used for machine 1 point
translation, say translating English to French? (Check all that apply.) *

It can be trained as a supervised learning problem.

It is strictly more powerful than a Convolutional Neural Network (CNN).

It is applicable when the input/output is a sequence (e.g., a sequence of words).

RNNs represent the recurrent process of Idea->Code->Experiment->Idea->....

10. Can a neural network learn & recall at the same time? * 1 point

YES

NO

38. Which of the following techniques can be used for normalization in text 1 point
mining? *

Stemming

Lemmatization

Stop Word Removal

Both A and B

1. Does pattern classification belongs to category of non-supervised 1 point

YES

NO

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 12/14
5/12/2020 MST-2_Machine Learning

37. Ability to learn how to do tasks based on the data given for training or 1 point
initial experience *

Self Organization

Adaptive Learning

Fault tolerance

Robustness

36. If calculation of update gate in GRU unit is close to 1, which of the 1 point
following would occur? *

Forgets the information for future time steps

Copies the information through many time steps

22. The training examples closest to the separating hyperplane are called 1 point
as _______ *

Training vectors

Test vectors

Support vectors

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 13/14
5/12/2020 MST-2_Machine Learning

17. Does backpropagaion learning is based on gradient descent along error 1 point

surface? *

YES

NO

cannot be said

it depends on gradient descent but not error surface

5. Learning is a? * 1 point

slow process

fast process

can be slow or fast in general

can’t say

Page 2 of 2 Back Submit

Never submit passwords through Google Forms.

This form was created inside of Acropolis Institute of Technology and Research. Report Abuse

 Forms

https://docs.google.com/forms/d/e/1FAIpQLSem15gzca-m2XzSWpRj2gW8lsTesjN6_NZtWjXjbFglUilUEw/formResponse 14/14