wayground logo

Free Printable Worksheets

NEW

Font size

S
M
L
XL
Worksheets

Deep Learning

Total questions: 52

Worksheet time: 37mins

Name
Class
Date
1.

Which of the following is NOT supervised learning?

a)

Decision Tree

b)

Linear Regression

c)

Naive Bayes

d)

Clustering

2.

What is the primary purpose of Back Propagation Networks (BPN)?

a)

a) Image classification

b)

b) Language translation

c)

c) Minimizing error in neural networks

d)

d) Generating random numbers

3.

What is the basic building block of a neural network?

a)

Layer

b)

Neuron

c)

Weight

d)

Edge

4.

What is the main goal of deep learning?

a)

Unsupervised learning

b)

Feature extraction

c)

Training neural networks to learn from data and make predictions

d)

Reinforcement learning

5.

What does the term "ReLU" stand for in the context of activation functions?

a)

Random Element Linear Unit

b)

Resilient Learning Encoder

c)

Recursive Leverage Engine

d)

Rectified Linear Unit

6.

Who am I? I am the most fundamental block of a single-layer neural network?

a)

Neuron

b)

Node

c)

Synapse

d)

Perceptron

7.

In deep learning, what does the term "backpropagation" refer to?

a)

Forward movement of data

b)

Learning from mistakes

c)

Rewinding the network

d)

Initialization of weights

8.

What is the purpose of an activation function in a neural network?

a)

Adjusting weights

b)

Reducing bias

c)

Introducing non-linearity

d)

Controlling iterations

9.

Which type of neural network is commonly used for image recognition tasks?

a)

Feedforward Neural Network (FNN)

b)

Recurrent Neural Network (RNN)

c)

Convolutional Neural Network (CNN)

d)

Artificial Neural Network (ANN)

10.

What role does the "bias" play in a neural network?

a)

Adjusting the overall influence of inputs

b)

Reducing computation time

c)

Controlling learning rate

d)

Initializing weights

11.

What is the primary purpose of the "loss function" in training a neural network?

a)

Initializing weights

b)

Evaluating model performance

c)

Reducing bias

d)

Controlling learning rate

12.

Who is the god father of AI?

a)

Alan Turing

b)

Geoffrey Hinton

c)

Elon Musk

d)

John McCarthy

13.

Which type of neural network is suitable for sequential data, such as time series or natural language processing?

a)

Feedforward Neural Network (FNN)

b)

Recurrent Neural Network (RNN)

c)

Convolutional Neural Network (CNN)

d)

Radial Basis Function Network (RBFN)

14.

Which layer of a neural network is responsible for making the final predictions?

a)

Input layer

b)

Hidden layer

c)

Output layer

d)

Convolutional layer

15.

In an ANN, what does the term "feedforward" refer to?

a)

Forward propagation of data

b)

Backward propagation of errors

c)

Initializing weights

d)

Adjusting the learning rate

16.

Who Am I? I influence the importance of input features in training a neural network.

a)

Weights

b)

Bias

c)

Activation Function

d)

Loss Function

17.

What is the output range of the sigmoid activation function?

a)

[-1, 1]

b)

[0, 1]

c)

[-∞, ∞]

d)

[0.5, 1]

18.

1. What is the primary purpose of a Convolutional Neural Network (CNN)?

a) Object detection

b) Image classification

c) Text generation

d) Reinforcement learning

a)

A

b)

B

c)

C

d)

D

19.

Which layer type is typically used to extract local features in a CNN?

a) Convolutional layer

b) Pooling layer

c) Fully connected layer

d) Activation layer

a)

A

b)

B

c)

C

d)

D

20.

What is the advantage of using convolutional layers in a CNN?

a) They can capture local spatial patterns in the input data

b) They can handle sequential data

c) They can generate synthetic data

d) They can handle variable-length inputs

a)

A

b)

B

c)

C

d)

D

21.

What is the purpose of the pooling layer in a CNN?

a) To reduce the spatial dimensions of the feature maps

b) To introduce non-linearity to the network

c) To adjust the weights and biases of the network

d) To compute the gradients for backpropagation

a)

A

b)

B

c)

C

d)

D

22.

Which activation function is commonly used in the convolutional layers of a CNN?

a) ReLU (Rectified Linear Unit)

b) Sigmoid

c) Tanh (Hyperbolic Tangent)

d) Softmax

a)

A

b)

B

c)

C

d)

D

23.

What is the purpose of the stride parameter in a convolutional layer?

a) To determine the size of the receptive field

b) To control the step size of the convolution operation

c) To adjust the learning rate during training

d) None of the above

a)

A

b)

B

c)

C

d)

D

24.

Which layer type is used to reduce the spatial dimensions in a CNN?

a) Convolutional layer

b) Pooling layer

c) Fully connected layer

d) Activation layer

a)

A

b)

B

c)

C

d)

D

25.

What is the purpose of data augmentation in CNN training?

a) To increase the number of layers in the network

b) To introduce noise and variations in the training data

c) To adjust the learning rate during training

d) None of the above

a)

A

b)

B

c)

C

d)

D

26.

What is the purpose of the learning rate in CNN training?

a) To control the step size of the parameter updates during optimization

b) To adjust the size of the filters in the convolutional layers

c) To increase the number of layers in the network

d) None of the above

a)

A

b)

B

c)

C

d)

D

27.

What is the purpose of the loss function in CNN training?

a) To measure the prediction error and guide the learning process

b) To initialize the weights and biases of the network

c) To adjust the learning rate during training

d) None of the above

a)

A

b)

B

c)

C

d)

D

28.

Which layer type is commonly used in CNNs to normalize the input data?

a) Convolutional layer

b) Pooling layer

c) Batch normalization layer

d) Activation layer

a)

A

b)

B

c)

C

d)

D

29.

Which layer type is responsible for parameter sharing in a CNN?

a) Convolutional layer

b) Pooling layer

c) Fully connected layer

d) Activation layer

a)

A

b)

B

c)

C

d)

D

30.

What is the primary advantage of using a CNN over a fully connected neural network for image processing tasks?

a) CNNs have a higher training speed

b) CNNs can handle sequential data

c) CNNs have a higher number of neurons

d) CNNs can capture local spatial patterns in the input data

a)

A

b)

B

c)

C

d)

D

31.

What is the purpose of dropout regularization in a CNN?

a) To increase the number of layers in the network

b) To adjust the learning rate during training

c) To randomly disable neurons during training to prevent overfitting

d) None of the above

a)

A

b)

B

c)

C

d)

D

32.

What is the purpose of the fully connected layers in a CNN?

a) To capture global patterns and make predictions

b) To reduce the spatial dimensions of the input data

c) To apply non-linear transformations to the feature maps

d) To initialize the weights and biases of the network

a)

A

b)

B

c)

C

d)

D

33.

Which of the following is NOT supervised learning?

a)

Decision Tree

b)

Linear Regression

c)

Naive Bayes

d)

Clustering

34.

A Machine Learning method that is concerned with how software agents should take actions in an environment is

a)

Reinforcement learning

b)

semi- supervised learning

c)

unsupervised learning

d)

supervised learning

35.

What is the objective of backpropagation algorithm?

a)

to develop learning algorithm for multilayer feedforward neural network

b)

to develop learning algorithm for single layer feedforward neural network

c)

to propagate the computed loss to compute gradient

d)

all the mentioned

36.

What are general limitations of back propagation rule? Pick the right choice from the given options

i) local minima problem

ii) slow convergence

iii) scaling

a)

only i

b)

both ii and i

c)

both ii and iii

d)

i, ii and iii

37.

Feedback networks are used for?

a)

auto association

b)

pattern storage

c)

both auto association & pattern storage

d)

pattern recognition

38.

Number of output cases depends on what factor?

a)

number of inputs

b)

number of distinct classes

c)

total number of classes

d)

none of the mentioned

39.

_________ is used to find local minima of the cost function

a)

stochastic gradient descent

b)

gradient descent

c)

linear regression

d)

logistic regression

40.

Cost function(J) of Linear Regression is the _______ value between predicted y value (predicted) and true y value (y)

a)

Mean

b)

Root mean square

c)

Median

d)

Mean square

41.

A = 1/(1 + e-x) is an equation representing which activation function?

a)

ReLU

b)

Sigmoid

c)

Leaky ReLu

d)

Tanh

42.

The algorithm creates a line or a hyperplane which separates the data into classes and also suitable for classification and regression

a)

K-means clustering

b)

Support vector machine

c)

Bayesian inference

d)

perceptron

43.

What steps can we take to prevent overfitting in a Neural Network?

a)

Data Augmentation

b)

Early Stopping

c)

Dropout

d)

Regularization

e)

All the mentioned

44.

Identify the activation function from the given diagram

a)

Sigmoid, ReLU

b)

ReLU, Leaky ReLU

c)

ReLU, Tanh

d)

TanH, ReLU

45.

___________uses the processing of the brain as a basis to develop algorithms that can be used to model complex patterns and prediction problems.

a)

ANN

b)

CNN

c)

RNN

d)

KNN

46.

How many possible layers can be there in deep neural network

a)

1

b)

≥ 50

c)

10

d)

no limit

47.

In VC dimensions of neural networks what does VC stand for?

a)

Vladimir Cherubim

b)

Vapnik Chervonenkis

c)

Victor Charlie

d)

Vanessa Carlton

48.

The number of nodes in the input layer is 10 and the hidden layer is 5. The maximum number of connections from the input layer to the hidden layer are

a)

50

b)

Less than 50

c)

More than 50

d)

It is an arbitrary value

49.

In a simple MLP model with 8 neurons in the input layer, 5 neurons in the hidden layer and 1 neuron in the output layer. What is the size of the weight matrices between hidden output layer and input hidden layer?

a)

[1 X 5] , [5 X 8]

b)

[8 X 5] , [ 1 X 5]

c)

[8 X 5] , [5 X 1]

d)

[5 x 1] , [8 X 5]

50.

Which of the following neural network training challenge can be solved using batch normalization?

a)

Overfitting

b)

Restrict activations to become too high or low

c)

Training is too slow

d)

All the mentioned

51.

For a binary classification problem, which of the following architecture would you choose?

a)

1

b)

2

c)

Any one of these

d)

None of these

52.

The red curve above denotes training accuracy with respect to each epoch in a deep learning algorithm. Both the green and blue curves denote validation accuracy.

a)

Green Curve

b)

Blue Curve

c)

Red Curve

d)

None