Scribd
Upload
13 views

Deep Learning 20CSE21_previous paper

Uploaded by

sunnyrocks815
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
13 views

Deep Learning 20CSE21_previous paper

Uploaded by

sunnyrocks815
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

Code No.

: 20CSE21

CHAITANYA BHARATHI INSTITUTE OF TECHNOLOGY (Autonomous)


B.E. (CSE) VII Sem (Main) Examination Nov – Dec 2023
Deep Learning
Time: 3 Hours Max Marks: 60
Note: Answer ALL questions from Part-A & Part –B (Internal Choice) at one place in the
same order
Part - A
(5Q X 3M = 15 Marks)
M CO BT
1 Obtain the output of the neuron Y for the network shown in figure using (3) 1 3
activation function as binary sigmoidal function.

2 Outline the Data Set augmentation. (3) 2 2


3 Illustrate the convolution operation. (3) 3 2
4 List some typical applications of RNN. (3) 5 2
5 What are the objectives of the Generator and Discriminator in GAN? (3) 6 1

Part – B
(5Q X 9M = 45 Marks)
M CO BT
6 (a) Explain how to learn Multilayer Networks using Gradient Descent (5) 1 2
Algorithm.
(b) Draw multi-layer feed forward network architecture and explain (4) 2 2
briefly.
(OR)
7 (a) Explain Perceptron learning algorithm. Differentiate between Gradient (5) 1 2
Descent and Stochastic Gradient Descent.
(b) Explain how Adam is used for training deep learning models in local (4) 2 2
minima.

8 (a) Describe auto encoders and how they help in dimensionality reduction. (5) 2 2
(b) What is model over fitting? Explain how to reduce model overfitting (4) 4 2
with either Dropout or Batch Normalization.
(OR)

Page 1 of 2
Code No.: 20CSE21

9 (a) With the help of Autoencoder architecture explain the process of (5) 2 3
feature generation.
(b) Explain the role of Batch Normalization in training a neural network. (4) 4 2

10 (a) An input image has been converted into a matrix of size 12 X 12 along (5) 3 3
with a filter of size 3 X 3 with a Stride of 1. Determine the size of the
convoluted matrix.
(b) Suppose you have 40 convolutional kernel of size 3x3 with no padding (4) 3 3
and stride 1 in the first layer of a convolutional neural network. You
pass an input of dimension 1024x1024x3 through this layer. What are
the dimensions of the data which the next layer will receive?
(OR)
11 (a) Why do we prefer Convolutional Neural Networks (CNN) over (5) 3 3
Artificial Neural Networks (ANN) for image data as input?
(b) Sketch the AlexNet architecture and explain its functionalities. (4) 3 3

12 (a) Explain the working of Gated Recurrent Unit. (5) 5 2


(b) Explain the difference between LSTM and GRU. (4) 5 2
(OR)
13 (a) Summarize the LSTM working mechanism. (4) 5 2
(b) Demonstrate the Transformers with the help of neat diagram. (5) 5 3

14 (a) How GAN Architecture be used with a neat diagram? (5) 6 3


(b) Describe training a Generative Adversarial Network. (4) 6 2
(OR)
15 (a) Explain the GAN Architecture’s operation and its difference from (5) 6 2
other models.
(b) Explain briefly Image-to-image Translation using Conditional-GANs. (4) 6 2

******

Page 2 of 2

You might also like