Code No: R2032422 R20 SET -1

III B. Tech II Semester Regular Examinations, July -2023

DEEP LEARNING
(Com. To CSE(AIML), CSE(AI), AIML)
Time: 3 hours Max. Marks: 70
Answer any FIVE Questions ONE Question from Each unit
All Questions Carry Equal Marks
*****
UNIT-I
1. a) Compare Deep Learning with Machine Learning. [7M]
b) Probability theory is a fundamental tool of many disciplines of science and [7M]
engineering. Justify.
(OR)
2. a) List and explain the four branches of Machine Learning. [7M]
b) Explain in detail about K-Fold validation and Overfitting and Underfitting. [7M]
UNIT-II
3. a) Explain how deep learning is incorporated into human language applications. [7M]
b) Explain the architecture of Back Propagation networks and Describe Back [7M]
Propagation algorithm for training.
(OR)
4. a) Explain in detail about Artificial Neural Networks. [7M]
b) Give a brief history of deep learning for Natural Language Processing. [7M]
UNIT-III
5. a) Explain the deep-learning software and hardware stack. [7M]
b) What are the pros and cons of running deep-learning jobs in the cloud. [7M]
(OR)
6. a) How to use neural networks to Classifying movie reviews as positive or [7M]
negative.
b) How to set up the deep learning workstations? Explain with example. [7M]
UNIT-IV
7. a) Write an example function for Convolution and Pooling operations and explain [7M]
in detail.
b) Describe neural network representation and applications of CNN suitable [7M]
example.
(OR)
8. a) Draw and explain the architecture of Recurrent Neural Networks. [7M]
b) What are the features of PyTorch? Implement CNN in PyTorch [7M]
UNIT-V
9. a) Explain Generative Adversial Networks-Working principle and applications. [7M]
b) Discuss in detail about Autoencoders and challenges in its implementation. [7M]
(OR)
10. a) Explain Boltzmann Machines- Working principle and applications.. [7M]
b) What do you mean by stacked recurrent models? Explain. [7M]

1o1

|''|'||||''|'''|||'|
 Code No: R2032422
R2031011
R2031351
R203135A
R203147A
R203147C
R203105O
P2031051 R20 SET
SET
RA--22

III B. Tech II Semester Regular Examinations, July -2023

DEEP LEARNING
(Com. To CSE(AIML), CSE(AI), AIML)
Time: 3 hours Max. Marks: 70
Answer any FIVE Questions ONE Question from Each unit
All Questions Carry Equal Marks
*****
UNIT-I
1. a) What is Deep Learning? Describe how deep learning is related to artificial [7M]
intelligence and machine learning.
b) List and explain the historical trends in Deep Learning. [7M]
(OR)
2. a) Explain the Formal evaluation procedures for machine learning models [7M]
b) How to create decision trees? Explain its role in construction of random [7M]
forests.
UNIT-II
3. a) What is Machine Vision and Biological visual system? Explain. [7M]
b) How to mimic the biological neuron structure as Artificial neural networks? [7M]
Explain with architecture and working principle.
(OR)
4. a) What is a cost and activation functions? Analyze these functions. [7M]
b) Describe about learning conditional statistics in gradient based learning. [7M]
UNIT-III
5. a) What is multiclass classification? Explain the algorithm for classifying student [7M]
grade as Poor, Satisfactory, Good, Very Good.
b) List the key features of Keras? Write two options for running Keras. [7M]
(OR)
6. a) How to use neural networks to Classifying news wires by topic. [7M]
b) Explain the relationship between the network, layers, loss function, and [7M]
optimizer.
UNIT-IV
7. a) Draw and explain the architecture of convolution neural networks. [7M]
b) Illustrate the process of building a deep neural network in PyTorch. [7M]
(OR)
8. a) Write an example function for Multichannel convolution operation and explain [7M]
in detail.
b) Explain how to compute the gradient in a Recurrent Neural Network. [7M]
UNIT-V
9. a) What is Deep Reinforcement Learning. Explain types of Deep Reinforcement [7M]
Learning.
b) Explain Restricted Boltzmann Machines and implementation challenges. [7M]
(OR)
10. a) Explain in detail about interactive applications of Deep Learning with respect [7M]
to NLP.
b) Explain about Deep Belief Networks. [7M]

1 of 1

|''|'||||''|'''|||'|
 Code No: R2032422 R20 SET -3

III B. Tech II Semester Regular Examinations, July -2023

DEEP LEARNING
(Com. To CSE(AIML), CSE(AI), AIML)
Time: 3 hours Max. Marks: 70
Answer any FIVE Questions ONE Question from Each unit
All Questions Carry Equal Marks
*****
UNIT-I
1. a) Comparison between Artificial Intelligence, Machine Learning and Deep [7M]
Learning?
b) Give an overview about Decision trees, random forests, and gradient [7M]
boosting machines?
(OR)
2. a) What is Machine Learning? What are the species of Machine Learning [7M]
techniques? Explain
b) Explain various techniques involved in Evaluating the performance of a [7M]
Machine Leaning Model?
UNIT-II
3. a) Discuss about Tensorflow Playground? [7M]
b) Differentiate the working of forward propagation with hot dog-detecting [7M]
network?
(OR)
4. a) Explain the working of softmax function in the classification? [7M]
b) Discuss about the working of backpropagation with neat sketch? [7M]
UNIT-III
5. a) What is the Anatomy of a neural network? explain building blocks of [7M]
deep learning?
b) Give an overview of Developing with Keras in Deep Learning? [7M]
(OR)
6. a) What is Reuters dataset? How to pre-process and build the model on [7M]
dataset?
b) What is Multi-class classification? How to handle the labels and loss? and [7M]
also explain the importance of Intermediate layers?
UNIT-IV
7. a) Explain the relative positioning of the major concepts in Neural [7M]
Networks?
b) How to Develop a Multichannel CNN Model for Text Classification? [7M]
Explain
(OR)
8. a) Draw and explain Schematic diagram of a recurrent neural network? [7M]
b) Explain the basic operations performed in PyTorch in CNN? [7M]
UNIT-V
9. a) Explain about one of Deep learning application Computer Vision? [7M]
b) What is autoencoder? Explain various Regularized Autoencoders? [7M]
(OR)
10. a) What Are Generative Adversarial Networks? Explain [7M]
b) Explain about Reinforcement Learning Framework? [7M]

1 of 1
|''|'||||''|'''|||'|
 Code No: R2032422 R20 SET -4

III B. Tech II Semester Regular Examinations, July -2023

DEEP LEARNING
(Com. To CSE(AIML), CSE(AI), AIML)
Time: 3 hours Max. Marks: 70
Answer any FIVE Questions ONE Question from Each unit
All Questions Carry Equal Marks
*****
UNIT-I
1. a) What is Deep Learning and explain how it works with neat sketch. [7M]
b) What is Kernel trick? Explain about Kernel methods in SVM? [7M]
(OR)
2. a) What is overfitting? Explain how to prevent overfit by Reducing the [7M]
network’s size?
b) What is weight regularization? Explain about L1 and L2 regularizations? [7M]
UNIT-II
3. a) Explain how deep learning can be used for natural language processing? [7M]
b) Write the algorithmic procedure to train the deep networks. [7M]
(OR)
4. a) How can you use MODEL GENERALIZATION for avoiding overfitting? [7M]
b) Write and explain Keras code for Deep neural networks? [7M]
UNIT-III
5. a) What are the ways to setup the deep learning workstation? Explain. [7M]
b) What is hypothesis space and explain the functionalities of Loss functions [7M]
and Optimizers?
(OR)
6. a) Explain about IMDB Dataset and build the network on top of dataset? [7M]
b) How to validate your model? Explain how to generate predictions on new [7M]
dataset?
UNIT-IV
7. a) What is Representation Learning? Explain the Methods of Representation [7M]
Learning
b) Explain about the convolutional layers in CNN. [7M]
(OR)
8. a) Explain the Implementation of RNN in Keras. [7M]
b) What is tensors? Why are PyTorch Tensors Important for ML and DL? [7M]
UNIT-V
9. a) Discuss about Denoising Autoencoders? [7M]
b) Give the importance and working of n-grams in Natural Language [7M]
processing?
(OR)
10. a) Discuss about Restricted Boltzmann Machines? [7M]
b) Explain about non-convolutional models to admit training of deep [7M]
architectures?

1 of 1

|''|'||||''|'''|||'|