Course Code: Course Title: Deep Learning TPC 3 2 4
Version No. 1.0
Course Pre-requisites/
Co-requisites
None
Anti-requisites (if any). None
1. To understand the Fundamentals of Deep Learning
2. To learn Neural Network Architectures
Objectives: 3. To Implement Deep Learning Models
4. To apply Deep Learning to Solve Problems
5. To analyze and Optimize Model Performance
Course Outcomes
Course
Outcomes Course Outcome Statement
Understand the characteristics of deep learning models that are useful to
CO1
solve real-world problems.
CO2 Understand different methodologies to create application using deep nets.
Identify and apply appropriate deep learning algorithms for analyzing the
CO3
data for a variety of problems.
CO4 Implement different deep learning algorithms
CO5 Design the test procedures to assess the efficacy of the developed model.
TOTAL HOURS OF INSTRUCTIONS: 45
A
Module No. 1 Introduction to neural networks and deep neural 10Hours
networks
Neural Networks Basics - Functions in Neural networks – Activation function, Loss
function - Function approximation - Classification and Clustering problems - Deep
networks basics - Shallow neural networks – Activation Functions – Gradient Descent –
Back Propagation – Deep Neural Networks – Forward and Back Propagation –
Parameters – Hyperparameters
Module No. 2 Deep Learning Architectures 9Hours
Machine Learning and Deep Learning, Representation Learning, Width and Depth of
Neural Networks, Activation Functions: RELU, LRELU, ERELU, Unsupervised
Training of Neural Networks, Restricted Boltzmann Machines, Auto Encoders, Deep
Learning Applications
Module No. 3 Convolutional Neural Networks 6 Hours
Foundations of Convolutional Neural Networks – CNN operations – Architecture –
Simple Convolution Network – Deep Convolutional Models – ResNet, AlexNet,
InceptionNet and others.
Module No. 4 Sequence Modelling – Recurrent And Recursive Nets 12 Hours
Recurrent Neural Networks - Bidirectional RNNs, Encoder, Decoder, Sequence-to-
Sequence Architectures, Deep Recurrent Networks, Auto encoders - Bidirectional
Encoder Representations from Transformers (BERT), Long-Term Dependencies - Echo
State Networks - Long Short-Term Memory and Other Gated RNNs - Optimization for
Long-Term Dependencies - Explicit Memory.
Module No. 5 Transfer Learning&Deep Reinforcement Learning 8Hours
Transfer learning Techniques, Variants of CNN: DenseNet, PixelNet, Deep
Reinforcement Learning – Q-Learning – Deep Q-Learning – Policy Gradients -
Advantage Actor Critic (A2C) and Asynchronous Advantage Actor Critic (A3C) – Model
based Reinforcement Learning – Challenges.
Text Books
1. Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow (3rd Edition)
by Aurélien Géron, O’Reilly Media, 2022.
2.Deep Learning with Python (2nd Edition) by François Chollet, Manning Publications,
2021.
3. Introduction to Deep Learning by Eugene Charniak, MIT Press, 2020.
References
1. Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow (3rd Edition)
by Aurélien Géron, O’Reilly Media, 2022.
2. Deep Learning with Python (2nd Edition) by François Chollet, Manning Publications,
2021.
3.Practical Deep Learning: A Python-Based Introduction by Ronald T. Kneusel, No
Starch Press, 2021.
List of Laboratory Experiments
Week 1:
1: Write a program to simulate a perceptron network for pattern classification and
function approximation.
Week 2:
2: Write a program to solve a XOR function using feed-forward neural network trained
using backpropagation algorithm.
Week 3:
3: Write a program to implement adaptive noise cancellation using ADALINE neural
network.
Week 4 & 5:
4& 5: Given the region to be de-fuzzified, write programs to discuss the various methods
that might be chosen.
Week 6:
6: Implementation of simple Over Current Relay using fuzzy logic.
Week 7:
7: Simulation and comparison of fuzzy PID controller with conventional PID controller
for a given plant.
Week 8:
8: Solve optimal relay coordination as a linear programming problem using Genetic
Algorithm.
Week 9 & 10:
9 & 10: Solve optimal relay coordination as a non-Linear programming problem using
Genetic algorithm.
Week 11 & 12: Solve economic load dispatch problem using Genetic algorithm.
11 & 12: Write a program to simulate a perceptron network for pattern classification and
function approximation.
Course Type Embedded Theory and Lab (ETL)
Theory 75%
Examination-1 15
Examination-2 15
Assignment/ Quiz 20
Mode of Evaluation
Final Assessment Test 50
Laboratory 25%
Prepared by Mr.vipin kumar
Recommended by the
Board of Studies on
Date of Approval by
the Academic Council
