10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Dashboard
/
Courses
/
SCHOOL OF COMPUTING
/
ODD SEMESTER
/
SCSA3015
/
General

/
SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 1

Not yet answered

Marked out of 1.00

 Which of the following is the first

fundamental step in image processing?

a. Filtration

b. Image Acquisition

c. Image Enhancement

d. Image Restoration

Clear my choice

Question 2

Not yet answered

Marked out of 1.00

Choose
the best deep learning methods used for object recognition

a.
Restricted
Boltzmann Machine

b.
 Convolutional
Network, And Stacked Auto Encoder

c.
 All of
the Above

d.
Deep
Belief Networks

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 1/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 3

Not yet answered

Marked out of 1.00

What is 'gradient' when we are talking about RNN?

a. It is how RNN calls it´s features

b. The most important step of RNN algorhitm

c. A parameter that can help you improve the algorhitm´s accuracy

d. A gradient is a partial derivative with respect to its inputs

Clear my choice

Question 4

Not yet answered

Marked out of 1.00

 In a perceptron learning

algorithm, what is the initial value of the weights before the algorithm starts
learning?

a.
All weights assigned values
specific to the application in hand

b.
All weights set to zero

c. All weights set to one

d. All weights assigned random

values

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 2/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 5

Not yet answered

Marked out of 1.00

If a robot is able to change its own trajectory as per

the external conditions, then the robot is considered as the__

a.
Non-Servo

b. Open Loop

c. Intelligent

d. Mobile

Clear my choice

Question 6

Not yet answered

Marked out of 1.00

A generative model to
generate raw audio signal is

a.  Word2audio

b.
 NLP

c.
 Word2Vec

d.  WaveNet

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 3/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 7

Not yet answered

Marked out of 1.00

What is a faceprint?

a. code which identifies a face

b.  distinctive curves in the

face

c.  a 2D image of the face

d. the number of nodal points

on the face

Clear my choice

Question 8

Not yet answered

Marked out of 1.00

Which of the following

methods is commonly used to reduce this gap?

a.
 Generative Adversarial
Networks

b.
 Sigmoid activation

c. RMSprop optimizer

d. Dropout

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 4/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 9

Not yet answered

Marked out of 1.00

What is the main challenge of NLP?

a. Handling Ambiguity of
Sentences
A.     Handling Ambiguity of
Sentences

b. Handling POS-Tagging

c. None of these

d.  Handling Tokenization

Clear my choice

Question 10

Not yet answered

Marked out of 1.00

The other RNN´s issue is called 'Vanishing Gradients'. What is that?

a. When the values of a gradient are too big and the model joins in a loop because of that.

b. When the values of a gradient are too big and the model stops learning or takes way too long because of that.

c. When the values of a gradient are too small and the model stops learning or takes way too long because of that.

d. When the values of a gradient are too small and the model joins in a loop because of that.

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 5/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 11

Not yet answered

Marked out of 1.00

Let us assume we
implement an AND function to a single neuron. Below is a tabular representation
of an AND function:

X1 X2 X1 AND X2

0 0 0

0 1 0

1 0 0

1 1 1

What would be the weights and bias? 

(Hint: For which values of w1, w2 and b does our neuron implement an AND
function?)

a.
Bias = -1.5, w1 = 1, w2 = 1

b. Bias = 1.5, w1 = 2, w2 = 2

c. None of these

d. Bias = 1, w1 = 1.5, w2 = 1.5

Clear my choice

Question 12

Not yet answered

Marked out of 1.00

Which of the following is a

non-iterative method to generate adversarial examples?

a. Fast Gradient Sign Method

b.  Adversarial Training

c.
 Non-Saturating Cost Method

d.
Input Optimization Method

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 6/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 13

Not yet answered

Marked out of 1.00

For an
image recognition problem (recognizing a cat in a photo), which architecture of
neural network would be better suited to solve the
problem?

a.  Recurrent
Neural network

b.
 Perceptron

c.
 Multi
Layer Perceptron

d.
 Convolutional Neural Network

Clear my choice

Question 14

Not yet answered

Marked out of 1.00

The difference between deep learning and machine

learning algorithms is that there is no need of feature engineering in machine
learning algorithms, whereas, it is recommended to do feature engineering first
and then apply deep learning.

Select one:
True

False

Question 15

Not yet answered

Marked out of 1.00

Which of the following

propositions are true about a CONV layer? (Check all that apply.)

a.
The total number of
parameters depends on the stride.

b.
The number of weights
depends on the depth of the input volume.

c.
The number of biases is
equal to the number of filters.

d. The total number of parameters depends on the padding

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 7/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 16

Not yet answered

Marked out of 1.00

Pick out the one in biological neuron that is

responsible for receiving signal from other neurons

a.
 Dendrite

b. Axon

c. Synapse

d. Soma

Clear my choice

Question 17

Not yet answered

Marked out of 1.00

Consider a Generative
Adversarial Network (GAN) which successfully produces images of apples. Which
of the following propositions is false?

a.
The generator can produce
unseen images of apples.

b.
The discriminator can be
used to classify images as apple vs. non-apple.

c.
The generator aims to learn
the distribution of apple images.

d.
 After training the GAN, the
discriminator loss eventually reaches a constant value.

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 8/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 18

Not yet answered

Marked out of 1.00

Given the input image with dimensions 125 ×

49, filter of size 5 × 5, padding P=2P=2 and stride S=2S=2, what is the dimension of the output
image of
this convolution operation?

a. None of these

b.
  63*25

c.    64*24

d.   64*26

Question 19

Not yet answered

Marked out of 1.00

Successful
applications of Machine Learning

a.
Learning to
recognize spoken words

b. All of the above

c. Learning to drive an autonomous vehicle

d. Learning to classify new astronomical structures

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 9/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 20

Not yet answered

Marked out of 1.00

A generative model to
generate raw audio signal is

a.
 Word2audio

b.
 NLP

c.
 WaveNet

d.
 Word2Vec

Clear my choice

Question 21

Not yet answered

Marked out of 1.00

One of the RNN´s issue is 'Exploding Gradients'. What is that?

a. When the algorithm assigns a stupidly high importance to the weights, because the better features

b. When the algorithm assigns a stupidly high importance to the weights, when your dataset is too big

c. When the algorithm assigns a stupidly high importance to the weights, when your data is too small

d. When the algorithm assigns a stupidly high importance to the weights, without much reason

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 10/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 22

Not yet answered

Marked out of 1.00

Which of the following is true about Batchnorm?

a. Batchnorm is another way of

performing dropout.

b. Batchnorm is a non-linear
transformation to center the dataset around the origin

c.
Batchnorm makes training
faster.

d.
In Batchnorm, the mean is
computed over the features.

Clear my choice

Question 23

Not yet answered

Marked out of 1.00

Consider a 2D filter
convolution over a 3D input. What would be the dimension of the output?

a. 3

b. 2

c. 4

d. 1

Clear my choice

Question 24

Not yet answered

Marked out of 1.00

________occurs when the gradients become too large due to back-propagation.

a. RNN

b. Vanishing Gradient

c. LSTM

d. Exploding Gradients

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 11/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 25

Not yet answered

Marked out of 1.00

The available ways to solve a problem of

state-space-search.

a. 4

b. 3

c. 1

d. 2

Clear my choice

Question 26

Not yet answered

Marked out of 1.00

Which of the following

activation functions can lead to vanishing gradients?

a. None of these

b. ReLU

c. Tanh

d. Leaky ReLU

Clear my choice

Question 27

Not yet answered

Marked out of 1.00

In which of the following applications can we use deep

learning to solve the problem?

a. Detection of exotic
particles

b. Protein structure
prediction

c. Prediction of
chemical reactions

d. All of the above

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 12/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 28

Not yet answered

Marked out of 1.00

Which of the following statements is true when

you use 1×1 convolutions in a CNN?

a. It suffers less overfitting due to small kernel size

b. All of the above

c. It can be used for feature pooling

d. It can help in dimensionality reduction

Clear my choice

Question 29

Not yet answered

Marked out of 1.00

RNN Stands for _______.

a. Recursive Neural Network

b. Removable Neural Network

c. Recurrent Neural Network

d. Recurring Neural Network

Clear my choice

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 13/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Question 30

Not yet answered

Marked out of 1.00

 Which of the following can

be used to evaluate the performance of linear regression model

a. ROC

b.
 Log loss

c.
 Mean square loss

d.
 Mean square loss

Clear my choice

◄ SCSA3015 Deep Learning Assignment for IVth Year(Batch 2019 - 2023)

Jump to...

SCSA3015 Deep Learning Syllabus ►

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 14/15
10/23/22, 6:20 PM SCSA3015 Deep Learning Quiz for IV year (Batch 2019 - 2023)

Sathyabama Learning Management System

💻📚 e-Resources

INFO
Facebook
Twitter

Instagram
YouTube
Sathyabama Staff Forum
Course Material

GET SOCIAL



Copyright © 2022

https://sathyabama.cognibot.in/mod/quiz/attempt.php?attempt=1449435&cmid=76519 15/15