Code No: E-5960/NIAICTE
B.E. CSE (Al&ML) VI Semester FACULTY OF ENGINEERING
(AICTE) (Main) (New) Examination, September /
October 20z3
Time: 3 Hours Subject: Deep Learning Technlques
Note: () First question is compulsory and Max. Marks: 70
remaining six answer any four questions from the
(i) Answer to eachquestions, Each questions carries 14 Marks.
question must be
order as they occur in the questionwritten
paper.
at one place only and in the same
(iii) Missing data, if any, may be
suitably assumed.
1. a) What is the promising future of Al?
b) What are the advantages of neural
networks?
c) Explain the role of hyper parametertuning in neural network training.
d) What are the top five use cases of
TensorFlow?
e) Describe the shape or dimension of input data for a recurrent
neüral network.
f) Discuss the importance of pooling layer in Convolutional Neural
9) Distinguish between biological and artificial neuron. Network.
2. a) Howdeep learning works? What are various appliations of deep
)Differentiate in between in Artificial Intelligence aFd learning?
Deep Learning.
3. a) What is Perceptron? Explain the working
ofPerceptron?
b) List and explain the various activation functions used in modeling of artificial neuron.
Also explain their suitability with respect to applications.
4. a) What is feed-forward
network.
network?Explin gradient based learning in a feed forward
b)How to build a neural network using
network in TensorFlôw? TensorFlow? How do you visualize a neural
5. a) With the helpÑf adlàgYam, explain basic building blocks of Convolutional Neural
Network architecture.
b) Discuss about Recursive Neural
Networks with neat diagram.
6. a) What are various
)Explain the use ofimplementations available for large-scale deep
deep learning in Natural learning.
Language Processing.
7. a) What are various steps
required for installation of TensorFlow? Discuss about various
TensorFlow Operations.
B) Explain about
backpropogation algorithm for training a neural network.
kt
Code No: F-13906/NIAICTE
FACULTY OF ENGINEERING
B.E. CSE (AI&ML) V -Semester (AlCTE) (Backlog) (New) Examination, February/ March 2024
Subject: Deep Learning Techniques
Time: 3 Hours
Max, Marks: 70
Note: () First question is compulsory and answer any four questions from the
remainingsix questions.Each question carries 14 Marks.
(0) Answer to each question must be written at one place only aîd in the same
1602
order as they occur in the question paper.
(i) Missing data, if any, may be suitably assumed.
1. a) What is Deep Learning?
b) Define Linear Perception.
c) Explain Autoencoders.
d) What is Gradient Descent? U
e) What is TensorFlow?
f) What is Unsupervised Pretrained Network?
9) Give two application of Deep Learning.
2. a) Explain the concept of Learning representation from data.
b) How Deep Learning differs from Machine Learning justify?
3. a) Define Neural Network and discuss its properties.
b) Explain Activation Function in detail.
4. a) Write Backpropagation Algorithm and explain.
b) Explain Delta rule in detail.
5. a) Explain CNN architecture in detail.
b)Write basic steps to load CNN on MNIST.
6. a)Write short notes on NLP and NLMs.
b) Explain Speech Recognition.
7. b)a)Explain in detail the work flow for training and evaluating a deep learning model.
Explain how to create and manipulate TensorFlow variables and demonstrate TensorFlow
operations.
Code No: F-13906/N/BLIAICTE
FACULTY
B.E. CSE (Al&ML) VI- Semester
OF ENGINEERING
(AlCTE) (Main &Backlog) (New) Examination,
August / September 2024
Time: 3 Hours
Subject: Deep Learning Techniques
Max. Marks:70
Note: (i) First question is compulsory and answer any four
remaining six questions. Each question carries 14 questions
Marks.
from the
(ii) Answer to each question must be written at one place
order as they occur in the question paper. only and in the same
(ii) Missing data, if any, may be suitably assumed.
1. a) What is Deep Learning?
b) What kind of challenges you could face when applying neural networks?kidlen lyo 2s ato
c) What is loss function and optimizers?
d) What Is Tensor Flow?
e) What is the delta rule?
f) Determine the shape of output matrix of an image of size 19 X1gthat usesa 3 padding,size 2,
stride size 2, and a 5 x 5 filter.
g) List out various latest examples of NLP.
2. å) How deep leaning works ? Explain briefly.
b) Explain the concept of Learning representation of data.
3. a) Explain the concept of deep feed forwárd neural networks.
b) Explain various activtion funcion used in:modeling artificial neuron? Also explain its suitability
with respect to applications.
4. a) Explain the process of gradient descent and its role in optimizing the weights and biases of a neural
network. How does it heip in impfoving the accuracy of a deep learning model?
5)Discuss about building Multilayer Model for MNIST in Tensorflow.
5. a) Explain about Convolutional Neural Network architecture.
b) Discuss about RecurrentNeural Network (RNN)?
6. a) Write about various implementations available for large-scale deep learning.
b) Explain the use of deep learning in speech recognition.
7.a) Explain back propagation algorithm for neuralnetwork training.
b)Explain the architecture of LSTM.
