Scribd
Upload
59 views2 pages

Deep Learning

The document outlines the examination structure for a Deep Learning course, detailing the units and questions that students must answer. Each unit consists of two questions, with students required to answer one from each unit, totaling five questions for a maximum of 70 marks. The topics cover various aspects of deep learning, including neural networks, regularization techniques, convolutional layers, Hopfield Nets, and challenges in natural language processing.

Uploaded by

ysjenissy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
59 views2 pages

Deep Learning

The document outlines the examination structure for a Deep Learning course, detailing the units and questions that students must answer. Each unit consists of two questions, with students required to answer one from each unit, totaling five questions for a maximum of 70 marks. The topics cover various aspects of deep learning, including neural networks, regularization techniques, convolutional layers, Hopfield Nets, and challenges in natural language processing.

Uploaded by

ysjenissy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

H.T.No.

Code No: CT3548 SRGEC-R20


IV B.Tech I Semester Regular/Supplementary Examinations, December 2024
DEEP LEARNING
(Computer Science and Engineering, Artificial Intelligence and Data Science & Artificial
Intelligence and Machine Learning)
Time: 3 Hours Max. Marks: 70
Note: Answer one question from each unit.
All questions carry equal marks.
5 × 14 = 70M
UNIT-I
1. a) What are some common applications of feedforward neural networks in various domains?
(6M)
b) Explain in detail about Artificial Neural Networks of Deep learning. (8M)
(OR)
2. a) How do multi-layer neural networks enable more complex learning and representation?
(7M)
b) What are the different types of layers commonly used in a multi-layer neural network?
(7M)
UNIT-II
3. a) What are regularization techniques, and how do they help in risk minimization? (8M)
b) Can pretrained models obtained through greedy layer-wise training be applied to different
tasks effectively? (6M)
(OR)
4. a) What are the different issues that may encounter when training deep networks? (7M)
b) How does the vanishing gradient problem affect the flow of gradients through deep
networks? (7M)
UNIT-III
5. a) What is a convolutional layer, and how does it apply filters to input data? (7M)
b) How do pooling layers reduce the spatial dimensions of feature maps? (7M)
(OR)
6. What are Long Short-Term Memory (LSTM) cells, and how do they address the vanishing
gradient problem in RNNs? (14M)
UNIT-IV
7. a) How are Hopfield Nets utilized in pattern recognition and associative memory? (8M)
b) Discuss the concept of Neural Networks and its relationship with Sigmoid Nets. (6M)

Page 1 of 2
(OR)
8. a) Discuss the concept of energy minimization in Hopfield Networks. (7M)
b) Describe the architecture and purpose of autoencoders in neural networks. (7M)
UNIT-V
9. a) Explain the concept of sparse coding in machine learning, and provide an example of a
real-world application where sparse coding has been effectively employed to solve a
problem. (6M)
b) Discuss the characteristics, benefits and drawbacks of Theano and Torch in deep learning
tools. (8M)
(OR)
10. a) Outline the primary challenges in natural language processing (NLP). Give examples of
NLP applications and their impact on various industries. (7M)
b) Give an example of a transfer learning application in computer vision. Describe the
process of fine-tuning a pre-trained model for a specific task. (7M)

*****

Page 2 of 2

You might also like