0% found this document useful (0 votes)
2 views3 pages

Deep Learning

The document outlines a course on Deep Learning offered at Devi Ahilya University, detailing its objectives, prerequisites, and course outcomes. It covers key topics such as CNNs, RNNs, GANs, and Auto-encoders, along with a list of recommended books and assignments for practical application. The course aims to equip students with the skills to apply deep learning techniques to real-world problems across various domains.

Uploaded by

Obaid Abdullah
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
0% found this document useful (0 votes)
2 views3 pages

Deep Learning

The document outlines a course on Deep Learning offered at Devi Ahilya University, detailing its objectives, prerequisites, and course outcomes. It covers key topics such as CNNs, RNNs, GANs, and Auto-encoders, along with a list of recommended books and assignments for practical application. The course aims to equip students with the skills to apply deep learning techniques to real-world problems across various domains.

Uploaded by

Obaid Abdullah
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

Devi Ahilya University, Indore, India IVYear B.E.

Computer Engineering
Institute of Engineering & Technology (FullTime)

SubjectCode&Name Instructions Hours per Credits


Week
7CERE1 L T P L T P Total
Deep Learning 3 1 2 3 1 1 5
Duration of Theory
Paper: 3 Hours

LearningObjectives:
● To understand the basic concepts and techniques of Deep Learning and the need of Deep
Learning techniques in real-world problems.
● To understand CNN algorithms and the way to evaluate performance of the
CNNarchitectures
● To apply RNN and LSTM to learn, predict and classify the real-world problems in the
paradigms of Deep Learning.
● To understand, learn and design GANs for the selected problems.
● To understand the concept of Auto-encoders and enhancing GANs using auto-encoders
Prerequisites:Machine Learning.

Course Outcomes (CO) and Program Outcomes (PO) Mapping


CO Course Outcome Program Outcomes
No. (PO)
CO1 Understand the basic concepts and techniques of Deep PO1, PO2, PO12
Learning and the need of Deep Learning techniques in
real-world problems.
CO2 Understand CNN algorithms and the way to evaluate PO2, PO4, PO5,
performance of the CNN architectures. PO10
CO3 Apply RNN and LSTM to learn, predict and classify the PO2, PO4, PO5
real-world problems in the paradigms of Deep Learning.
CO4 Understand, learn and design GANs for the selected PO1, PO2, PO4,
problems. PO5, PO9, PO10,
PO11, PO12
CO5 Understand the concept of Auto-encoders and enhancing PO3, PO4, PO9,
GANs using auto-encoders. PO12

CO-PO Relationship Matrix


CO PO- PO- PO- PO- PO- PO- PO- PO- PO- PO- PO- PO-
1 2 3 4 5 6 7 8 9 10 11 12
CO1 3 2 - - - - - - - - - 2
CO2 - 2 - 2 3 - - - - 2 - -
CO3 - 3 - 3 3 - - - - - - -
CO4 3 3 - 3 3 - - - 2 2 2 3
CO5 - 2 3 3 - - - - 3 - - 3
COURSE OF CONTENTS
UNIT-I: INTRODUCTION TO DEEP LEARNING: Historical Trends in Deep Learning,
Why DL is Growing, Artificial Neural Network, Single/Multiple Layer Perceptron, Feed
Forward Network, Deep Feed-forward networks, Stochastic Gradient –Based learning, Back-
Propagation, Deep learning frameworks and libraries (e.g., TensorFlow/Keras, PyTorch).
UNIT-II: CONVOLUTION NEURAL NETWORK (CNN): Introduction to CNNs and
their applications in computer vision, CNN basic architecture, Activation functions-sigmoid,
tanh, ReLU, Softmax layer, Types of pooling layers, Training of CNN in TensorFlow, various
popular CNN architectures: VGG, GoogleNet, ResNet etc, Dropout, Normalization, Data
augmentation
UNIT-III: RECURRENT NEURAL NETWORK (RNN): Introduction to RNNs and their
applications in sequential data analysis, Backpropagation through time(BPTT), Vanishing
Gradient Problem, gradient clipping Long Short TermMemory (LSTM) Networks, Gated
Recurrent Units, Bidirectional LSTMs, Bidirectional RNNs.
UNIT-IV: GENERATIVE ADVERSARIAL NETWORKS (GANS): Generative models,
Concept and principles of GANs, Architecture of GANs (generator and discriminator
networks), Comparison between discriminative and generative models, Generative Adversarial
Networks(GANs), Applications of GANs
UNIT-V: AUTO-ENCODERS: Auto-encoders, Architecture and components of auto-
encoders (encoder and decoder), Training an auto-encoder for data compression and
reconstruction, Relationship between Autoencoders and GANs, Hybrid Models: Encoder-
Decoder GANs.

RECOMMENDED BOOKS:
1. Deep Learning : An MIT Press Book by Ian Goodfellow and Yoshua Bengio
Aaron Courville.
2. Michael Nielson,Neural Networks and Deep Learning,Determination Press,2015.
3. Satish kumar,Neural networks:A classroom Approach,Tata McGraw-Hill
Education.

LearningOutcomes:
After completing this course, the students will gain an understanding of the computational
behavior of the area of deep learning in order to apply it in various complex problems
including NLP, Image processing, game playing, robot learning, driverless cars etc.

List of Assignment in Deep Learning Lab:


.
● Build an Artificial Neural Network by implementing the Backpropagation algorithm and
test the same using appropriate data sets. Vary the activation functions used and compare
the results
● Build a Deep Feed Forward ANN by implementing the Backpropagation algorithm and
test the same using appropriate data sets. Use the number of hidden layers >=4.
● Design and implement an Image classification model to classify a dataset of images using
Deep Feed Forward NN. Record the accuracy corresponding to the number of epochs..
● Design and implement a CNN model (with 2 layers of convolutions) to classify multi
category image datasets. Record the accuracy corresponding to the number of epochs.
● Design and implement a CNN model (with 4+ layers of convolutions) to classify multi
category image datasets. Use any suitable dataset. Set the No. of Epoch as 5, 10 and 20.
Make the necessary changes whenever required. Record the accuracy corresponding to the
number of epochs. Record the time required to run the program, using CPU as well as
using GPU in Colab.
● Use the concept of Data Augmentation to increase the data size from a single image.
● Implement RNN for sentiment analysis on movie reviews
● Implement Bidirectional LSTM for sentiment analysis on movie reviews.
● Implement Generative Adversarial Networks to generate realistic Images. Use any human
face datasets.
● Implement Auto encoders for image denoising on any suitable dataset.

You might also like