Suppose that you are training a neural network for classification, but you notice that the training loss is much lower than the validation loss. Which of the following can be used to address the issue (select all that apply)?

Use a network with fewer layers

Increase L2 regularization weight

Increase the size of each hidden layer

Backpropagation works by first calculating the gradient of ___ and then propagating it backwards.

Sum of squared error with respect to inputs

Sum of squared error with respect to weights

Sum of squared error with respect to outputs

Given above is a description of a neural network. When does a neural network model become a deep learning model?

When you add more hidden layers and increase depth of neural network

When there is higher dimensionality of data

When the problem is an image recognition problem

When performing regression or classification, which of the following is the correct way to preprocess the data?

Normalize the data → PCA → training

PCA → normalize PCA output → training

Normalize the data → PCA → normalize PCA output → training

Which architecture of neural network would be better suited to solve the problem?

Convolutional neural network followed by recurrent units

End-to-End fully connected neural network

Perplexity is a commonly used evaluation technique when applying deep learning for NLP tasks. Which of the following statement is correct?

Lower the perplexity the better

Higher the perplexity the better

What is generally the sequence followed when building a neural network architecture for semantic segmentation for image?

Convolutional network on input and deconvolutional network on output

Deconvolutional network on input and convolutional network on output

Which of the following is TRUE about Perceptrons?

It is an artificial neuron and is the fundamental unit of a neural network

It is also called single layer neural network or search single layer binary linear classifier

A neural network is a composition of perceptrons connected in different ways

Which of the following is FALSE about Perceptrons?

Perceptron can learn both linearly and non-linearly separable functions

It cannot implement XOR gate as it cannot be classified by a linear separator

It can easily implement OR and AND gates as these are linearly separable

Which of the following is FALSE about Hidden Layers in Neural Networks?

Increasing the number of hidden layers always leads to higher accuracy

Abstract representation of the training data is stored in the hidden layers

Feature extraction happens at the hidden layers

Increasing the number of hidden layers above a certain point may lead to overfitting

Which of the following is FALSE about Deep Neural Networks?

These have less generalization capabilities as compared to shallow networks

These are computationally more complex as compared to shallow networks

These may suffer from overfitting problem

These may suffer from vanishing gradients problem

Which of the following is TRUE about Shallow Neural Networks?

These are computationally more complex as compared to deep neural networks

These have better generalization capabilities as compared to deep neural networks

These are better in learning and extracting features as compared to deep neural networks

Which of the following is a type of neural network?

CNN (Convolutional Neural Network)

Which of the following is a type of neural network?

LSTM (Long Short-Term Memory Networks)

Restricted Boltzmann Machines (RBM)

Which of the following is FALSE about Radial Basis Function Neural Network?

It resembles to RNNs which have feedback loops

It uses radial basis function as activation function

While outputting, it considers the distance of a point with respect to the center

What is the purpose of an activation function?

To decide whether a neuron will fire or not

To increase the depth of a neural network

To create connectivity among hidden layers

Which of the following is FALSE about step activation function?

It is also called Threshold activation function

Which of the following is FALSE about step activation function?

It is a differentiable activation function

Derivative of a step function is always zero

Which of the following is FALSE about sigmoid and tanh activation function?

Output of sigmoid ranges from 1 to 1 while output of tanh ranges from 0 to 1

Both are non-linear activation functions

Output of both sigmoid and tanh is smooth, continuous and differentiable

Which of the following is FALSE about sigmoid and tanh activation function?

Sigmoid is zero centered as its values range from 1 to 1

These cannot blow up the activations unlike ReLU

Both functions output a probability value instead of discrete values like 0 and 1

Which of the following is FALSE about sigmoid and tanh activation function?

These do not suffer from vanishing and exploding gradient problems unlike ReLU

These involve computing the exponential (which may be expensive) unlike ReLU

These are non-linear in nature like ReLU

ReLU activation function outputs zero when:

Input is less than or equal to zero

Input is greater than or equal to zero

ReLU outputs same value as input when:

Input is greater than or equal to zero

Input lies between zero and one

Which of the following is FALSE about ReLU activation function?

ReLU is preferred over sigmoid and tanh for output layer in a neural network

It does not require exponent calculation unlike sigmoid and tanh

It outputs o, if the input is less than or equal to o, otherwise output will be same as the input

Which of the following is FALSE about Dying ReLU?

Dying ReLU problem occurs when learning rate is too low

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

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

Which of the following is TRUE about Leaking ReLU?

Leaky ReLU is the most common method to alleviate a dying ReLU

Leaking ReLU adds a slight slope in the negative range to prevent dying ReLU issue

Parametric ReLU (PRELU) is a type of leaky ReLU that uses parameterized slope value

Deep Learning Quiz

Authored by Đặng 2C-20CACN

Computers

University

Used 4+ times

AI Actions

Add similar questions

Adjust reading levels

Convert to real-world scenario

Translate activity

More...

Content View

Student View

72 questions

Show all answers

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

While reconstructing images from embeddings, if we consider deeper layers of the network, how is the image affected?

The reconstructed image is blur

A more realistic copy of the original image is reconstructed

The reconstructed image is sharpened

The reconstructed image becomes abstract

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

Which of the following is true for content and style of image w.r.t. Deep Art?

Content should match and Style shouldn‟t match

Content shouldn‟t match and Style should match

Both Content and Style should match

Both Content and Style shouldn‟t match

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

It is very easy to fool a CNN. Which of the following statements doesn‟t support the above argument?

The CNN draws very stringent/strict boundaries for classification.

The decision boundaries are loosely drawn.

The input dimension for a certain class is very very large.

None

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

Which of the following includes major tasks of NLP?

Discourse Analysis

Automatic Summarization

Machine Translation

All of the above

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

Which of the following is used to mapping sentence plan into sentence structure?

Text planning

Sentence planning

Text Realization

None of the Above

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

Given a sound clip of a person or people speaking, determine the textual representation of the speech.

Text-to-speech

Speech-to-text

All of the mentioned

None of the mentioned

MULTIPLE CHOICE QUESTION

30 sec • 1 pt

is the step in which an input sentence is converted into a hierarchical structure that corresponds to the units of meaning in the sentence.

Syntactic Processing

Semantic Processing

Graph Processing

All of the mentioned

Access all questions and much more by creating a free account

Create resources

Host any resource

Get auto-graded reports

Continue with Google

Continue with Email

Continue with Classlink

Continue with Clever

or continue with

Microsoft

Apple

Others

Already have an account?

Similar Resources on Wayground

77 questions

SC-900 (81-167)

Quiz

•

University

72 questions

IAS Midterm Reviewer

Quiz

•

University

72 questions

Data Warehousing

Quiz

•

University

68 questions

Salesforce Test 1 For Ranking

Quiz

•

University

71 questions

PROTOCOLOS-2-ESPAÑOL

Quiz

•

University - Professi...

75 questions

Career Development Semester Exam 2025

Quiz

•

8th Grade - University

70 questions

Quiz 2 - JK3 - 2TK

Quiz

•

University

70 questions

Soal UTS Pemrograman Mobile

Quiz

•

University

Popular Resources on Wayground

15 questions

Fractions on a Number Line

Quiz

•

3rd Grade

20 questions

Equivalent Fractions

Quiz

•

3rd Grade

25 questions

Multiplication Facts

Quiz

•

5th Grade

54 questions

Analyzing Line Graphs & Tables

Quiz

•

4th Grade

$fractions$

22 questions

fractions

Quiz

•

3rd Grade

20 questions

Main Idea and Details

Quiz

•

5th Grade

20 questions

Context Clues

Quiz

•

6th Grade

15 questions

Equivalent Fractions

Quiz

•

4th Grade

Discover more resources for Computers

20 questions

CompTIA Network+ - Ports and Protocols

Quiz

•

University

Deep Learning Quiz

While reconstructing images from embeddings, if we consider deeper layers of the network, how is the image affected?

Which of the following is true for content and style of image w.r.t. Deep Art?

It is very easy to fool a CNN. Which of the following statements doesn‟t support the above argument?

Which of the following includes major tasks of NLP?

Which of the following is used to mapping sentence plan into sentence structure?

Given a sound clip of a person or people speaking, determine the textual representation of the speech.

is the step in which an input sentence is converted into a hierarchical structure that corresponds to the units of meaning in the sentence.

Suppose that you are training a neural network for classification, but you notice that the training loss is much lower than the validation loss. Which of the following can be used to address the issue (select all that apply)?

Which provides a framework for studying object recognition?

A recurrent neural network can be unfolded into a full-connected neural network with infinite length

Access all questions and much more by creating a free account

Similar Resources on Wayground

Popular Resources on Wayground

Discover more resources for Computers