Deep Learning

B.Tech (B)
(2023-24)(SEM-I)
Ms. Punam R. Patil
Vision of the Department:
To provide prominent computer engineering education with socio-moral values.

Mission of the Department:

1. To provide state-of-the-art ICT based teaching-learning process.
2. To groom the students to become professionally sound computer engineers to
meet growing needs of industry and society.
3. To make the students responsible human being by inculcating ethical values

Program Educational Objectives(PEOs)

PEO1: To provide the foundation of lifelong learning skills for advancing their
careers being a professional entrepreneur and leader
PEO2: To develop computer professionals to fulfill Industry expectations
PEO3: To foster ethical and social values to be socially responsible human being.

2
 Deep Learning (PECO7031T)

Teaching Scheme Examination Scheme

Lectures : 03 Hrs./week Term Test : 15 Marks
Credits : 03 Teacher Assessment : 20 Marks
End Sem Exam : 65 Marks
Total Marks : 100 Marks

3
Course Objectives:
Prerequisite: Artificial Intelligence, Machine Learning.

Course Objectives:
• To understand Hyper parameter Tuning.
• To explore Deep Learning Techniques with different learning
strategies.
• To design Deep Learning Models for real time applications.

4
Course Outcomes

5
Blooms Taxonomy

6
7
8
9
Unit-I: Introduction to Deep Learning 4 Hrs

• Overview of Neural Network, Deep learning and human

brain, Why is Deep Learning taking off?,
• Deep Learning applications.
• Overview of Tools: Torch, TensorFlow, Keras.

10
Unit II : Hyperparameter Tuning, Batch Normalization
05 Hrs

• Tuning Process, Using an Appropriate Scale to pick

Hyperparameters
• Hyperparameters Tuning in Practice: Pandas vs. Caviar,
Normalizing Activations in a Network, Fitting Batch Norm into a
Neural
• Network, why does Batch Norm work, Batch Norm at Test Time.

11
Unit III: Convolutional Neural Network
09 Hrs
• Introduction to CNNs, Kernel filter, Principles behind CNNs,
Multiple Filters, CNN applications.
• ConvNet Architectures
• Discussions on famous convnet architectures: AlexNet, VGG,
GoogLeNet, ResNet.

12
Unit IV: Recurrent Neural Networks
10 Hrs
• Introduction to Sequence Models and RNNs, Recurrent Neural
Network Model, Backpropagation
• Through Time, Different Types of RNNs: Unfolded RNNs, Seq2Seq
RNNs, Long Short-Term Memory (LSTM), Bidirectional RNN,
Vanishing Gradients with RNNs, Gated Recurrent Unit (GRU),
• RNN applications.

13
Unit V: Adversarial Networks 10 Hrs

• Introduction to adversarial Networks

• Auto encoders (standard, denoising, contractive, etc.)
• Vibrational Auto encoders
• Generative Adversarial Networks
• Applications of Adversarial Networks.

14
Unit VI: Deep Learning Case Studies
04 Hrs
• Image Processing
• Natural Language Processing
• Speech Recognition
• Video Analytics

15
Text Books:
1. Goodfellow I., Bengio Y., and Courville A., “Deep Learning”, MIT Press, 2016.
2. Umberto Michelucci, “Advanced Applied Deep Learning: Convolutional Neural Networks and Object Detection”,
2019.
3. Michael Nielsen (Goodreads Author), “Neural Networks and Deep Learning”, 2015.
4. Gulli and Kapoor, “TensorFlow 1.x Deep Learning Cookbook”, Packt Publishing, 2017.

Reference Books:
1. Yegnanarayana, B., “Artificial Neural Networks”, PHI Learning Pvt. Ltd, 2009.
2. Satish Kumar, “Neural Networks: A Classroom Approach”, Tata McGraw-Hill Education, 2004.
3. Rau´l Rojas, “Neural Networks: A Systematic Introduction”, 1996.
4. David Foster, “Generative Deep Learning: Teaching Machines to Paint, Write, Compose, and Play”, O’Relly, 2019.
5. Maxim Lapan, “Deep Reinforcement Learning HandsOn: Apply modern RL methods, with deep Q-networks, value
iteration, policy gradients, TRPO, AlphaGo Zero and more”, Packt Publishing, 2018.
6. Santanu Pattanaya K., “Pro Deep Learning with TensorFlow A Mathematical Approach to Advanced Artificial
Intelligence in Python”, APress, 2017.

16
Digital Material:

1. NPTEL:
Deep Learning, By Prof. Prabir Kumar Biswas, IIT Kharagpur.
https://onlinecourses.nptel.ac.in/noc22 cs22/preview
2. Coursera:
Deep Learning Specilization, By DeepLearning.AI
https://www.coursera.org/specializations/deep-learning#courses

17
18
List of Laboratory Experiments
• Suggested List of Experiments:
1. Building own Neural Network from scratch.
2. To implement EBPTA algorithm.
3. Understanding ANN using Tensor Flow.
4. Visualizing Convolutional Neural Network using Tensor Flow with
Keras Data.
5. Object detection using RNN using Tensor Flow.

19
6. Students are supposed to complete any one mini project not limited to
following list of projects.
• Sequence Prediction
• Object Detection
• Traffic Sign Classification
• Automatic Music Generation
• Music Genre Classification
• Text Summarizer
• Gender and Age Detection Using Voice
• Chatbot Using Deep Learning
• Neural Style Transfer
• Face Aging
• Driver Drowsiness Detection
• Language Translator
• Image Reconstruction
20
What is learning?

21
Can You Recognize these Pictures ?

• If Yes, How do you Recognize it?

Image Source: Internet 22

Origin of Machine Learning?
…..Lies in very early efforts of understanding Intelligence.

• What is Intelligence?
• It can be defined as the ability to comprehend; to understand and
profit from experience.
• Capability to acquire and Apply Knowledge.

Image Source: Internet 23

Learning?
2300 Years ago….
• Plato (427-347 BC )
• The concept of Abstract Ideas
are known to us a priori, through
a Mystic connection with world.
• He concluded that ability to think
is found in a priori knowledge of
the concepts.

Image Source: Internet 24

Learning? Plato’s
Pupil…
• Aristotle (384-322 BC)
• Criticized his Teacher’s Theory
as it is not taking into account
the important aspect
--- An ability to Learn or Adapt to changing
world.

Image Source: Internet 25

 Deep learning vs. machine learning
• Machine learning and deep learning are both types of AI. In short, machine
learning is AI that can automatically adapt with minimal human interference.
• Deep learning is a subset of machine learning that uses artificial neural networks to
mimic the learning process of the human brain.

https://www.coursera.org/articles/ai-vs-deep-learning-vs-machine-learning-beginners-guide 26
https://www.coursera.org/articles/ai-vs-deep-learning-vs-machine-learning-beginners-guide 27
 What Is Deep Learning?
• Deep learning is a subset of machine learning that
uses algorithms inspired by the structure and
function of the human brain, known as artificial
neural networks, to learn from large sets of data.
• That means, Deep learning is a machine learning
subset that makes computers do what comes
naturally to humans: learn by example.
• Deep learning models can automatically identify
patterns and features in data without being
explicitly programmed to do so.

In broad terms, deep learning is a subset of machine learning, and machine

learning is a subset of artificial intelligence.
28
 What Is Deep Learning?
• Think of deep learning as an evolution of machine learning.
• Deep learning is a machine learning technique that layers
algorithms and computing units—or neurons—into what is
called an artificial neural network.
• These deep neural networks take inspiration from the structure of the
human brain.
• Data passes through this web of interconnected algorithms in a non-
linear fashion, much like how our brains process information.
Today Deep learning has become one of the most popular and visible areas of machine learning, due to its
success in a variety of applications, such as computer vision, natural language processing, and
Reinforcement learning.
29
https://www.geeksforgeeks.org/introduction-deep-learning/?ref=lbp
Biological Neuron
• To understand how an artificial neuron works, we should first understand how a
biological neuron works.
• A nerve cell “neuron” is a special biological cell that processes information.
According to an estimation, there are huge number of neurons, approximately 1011
with numerous interconnections, approximately 1015.

30
Image Source: https://towardsdatascience.com/deep-learning-versus-biological-neurons-floating-point-numbers-spikes-and-neurotransmitters-6eebfa3390e9
Biological Neural Network

31
 Working of a Biological Neuron
• As shown in the above diagram, a typical neuron consists of the following 4 parts
with the help of which we can explain its working,

1. Dendrites −They are tree-like branches, responsible for receiving the information
from other neurons it is connected to. In other sense, we can say that they are like
the ears of neuron.
2. Soma − It is the cell body of the neuron and is responsible for processing of
information, they have received from dendrites.
3. Axon − It is just like a cable through which neurons send the information.
4. Synapses − It is the connection between the axon and other neuron dendrites.

32
BNN v/s ANN

Biological Neural Artificial Neural

Network BNN Network ANN

Soma Node
Dendrites Input
Synapse Weights or Interconnections
Axon Output

33
 What Is a Neural Network?
• A neural network is a method in artificial intelligence that teaches computers to
process data in a way that is inspired by the human brain mimicking the way that
biological neurons signal to one another.
• It is a type of machine learning process, called deep learning, that uses
interconnected nodes or neurons in a layered structure that resembles the human
brain.
• It creates an adaptive system that computers use to learn from their mistakes and
improve continuously.
• Thus, artificial neural networks attempt to solve complicated problems, like
summarizing documents or recognizing faces, with greater accuracy.

34
https://aws.amazon.com/what-is/neural-network/#:~:text=It%20is%20a%20type%20of,their%20mistakes%20and%20improve%20continuously
What Is a Neural Network?
• As a result, it’s worth noting that the “deep” in deep learning is just
referring to the depth of layers in a neural network.
• A neural network that consists of more than three layers—which
would be inclusive of the inputs and the output—can be
considered a deep learning algorithm.
• A neural network that only has two or three layers is just a basic
neural network.

35
https://www.ibm.com/topics/neural-networks
 Why are neural networks important?
• Neural networks can help computers make intelligent decisions with limited human
assistance. This is because they can learn and model the relationships between
input and output data that are nonlinear and complex. For instance, they can do the
following tasks.
Make generalizations and inferences
• Neural networks can comprehend unstructured data and make general
observations without explicit training. For instance, they can recognize that two
different input sentences have a similar meaning:
• Can you tell me how to make the payment?
• How do I transfer money?
Example: A neural network would know that both sentences mean the same thing. Or
it would be able to broadly recognize that Baxter Road is a place, but Baxter Smith is a
person’s name.
https://aws.amazon.com/what-is/neural-network/#:~:text=It%20is%20a%20type%20of,their%20mistakes%20and%20improve%20continuously 36
Why is Deep Learning taking off?

37
https://towardsdatascience.com/why-deep-learning-is-needed-over-traditional-machine-learning-1b6a99177063
• A big advantage with deep learning, and a key part in understanding
why it’s becoming popular, is that it’s powered by massive amounts of
data. The “Big Data Era” of technology will provide huge amounts of
opportunities for new innovations in deep learning.
• As per Andrew Ng, the chief scientist of China’s major search engine
Baidu and one of the leaders of the Google Brain Project, “The analogy
to deep learning is that the rocket engine is the deep learning
models and the fuel is the huge amounts of data we can feed to these
algorithms.”
• Deep Learning requires high-end machines contrary to traditional
Machine Learning algorithms. GPU has become a integral part now to
execute any Deep Learning algorithm.

38
https://towardsdatascience.com/why-deep-learning-is-needed-over-traditional-machine-learning-1b6a99177063
• Deep Learning techniques tend to solve the problem end to end, where
as Machine learning techniques need the problem statements to break
down to different parts to be solved first and then their results to be
combine at final stage.
39
https://towardsdatascience.com/why-deep-learning-is-needed-over-traditional-machine-learning-1b6a99177063
40
• By comparing two picture above, we see that traditional algorithms reach to a
threshold on performance. However, NN always works better with more data.
So you can get better performance as long as you collecting more and more
data, without changing the algorithm itself.
• The three factor are: algorithm, data and computation.
1. Algorithm improved a lot in NN area, such the introduce of ReLU activator.
2. In the internet era, we have more data then ever before;
3. the computational power has grown exponentially in recent years, thanks to
the cloud computing and GPU development.
These factors combined with the key differentiation make NN the
dominant algorithm in machine learning nowadays.

https://medium.com/@shaoliang.jia/why-is-deep-learning-taking-off-8244175929e7 41
Applications of Deep Learning
• Computer Vision
• Natural Language Processing
• Healthcare
• Finance
• Agriculture
• Cybersecurity
• Self Driving Cars ….

https://www.dataquest.io/blog/6-most-common-deep-learning-applications/ 42
 Applications of Deep Learning

43
https://www.pickl.ai/blog/top-applications-of-deep-learning-you-should-know/
Deep Learning and Computer Vision
• Most artificial intelligence applications have been developed on the shoulders of
computer vision, and deep learning has made significant advancements in computer
vision, including object detection, image classification, and image segmentation.
• For example, deep learning algorithms can be trained to detect specific objects
within an image, such as people or objects. This technology is essential for
applications including autonomous driving and robotics, wherein the vehicles or
robots must be able to detect and respond to different objects in their environment
— or in healthcare to analyze medical images.
• Deep learning is also being used to create more immersive virtual and augmented
reality experiences. By training neural networks on large datasets of images and
videos, developers can create more lifelike virtual environments that are responsive
to user inputs. This technology is being used in various applications, including video
games, training simulations, and architectural design — among other things.
• Overall, deep learning transforms computer vision, enabling machines to interpret
and understand visual information in a way that was once impossible. 44
 Deep Learning and Natural Language Processing
• Natural language processing (NLP) is the field of study that focuses on the interaction
between human language and computers. Deep learning has transformed the field of
NLP, enabling computers to interpret and respond to human language. One of the key
applications of deep learning in NLP is machine translation.
• Deep learning models can be trained on large datasets of human language to translate
text from one language to another accurately. This technology has a wide range of
applications, including website localization, document translation, and language
learning tools.
• Another important application of deep learning in NLP is sentiment analysis.
Sentiment analysis involves analyzing text to determine the emotional tone or
sentiment expressed by the author. Deep learning models can be used to classify text
accurately as positive, negative, or neutral, allowing businesses to gain insights into
customer opinions and feedback on their products or services. This can help
businesses make more informed decisions and improve customer satisfaction.
• Overall, deep learning has opened up a whole new world of possibilities in the field of NLP,
enabling us to understand and interact with human language better than ever before. 45
Deep Learning and Healthcare
• Deep learning improves medical imaging in healthcare, including CT and MRI scans, and it allows
specialists to develop more accurate and personalized treatment plans.
For example, deep learning algorithms can be used to analyze medical images, such as X-rays and
MRIs, to detect abnormalities that might not be visible to the naked eye.
Deep learning is also used to develop personalized medicine, where treatments are tailored to an
individual’s genetic makeup.
• Deep learning algorithms can analyze vast amounts of medical data to identify patterns and trends
that would be difficult for human physicians to detect. For example, deep learning algorithms can
identify early signs of cancer in mammograms, they can detect the presence of blood clots in brain
scans, and they can diagnose degenerative diseases in spinal cord images. This has led to the
development of new diagnostic tools and treatment options for conditions such as cancer, heart
disease, and neurological disorders.
• Overall, deep learning has the potential to transform healthcare with applications that can
improve diagnostics, treatment, and patient outcomes. As the technology continues to
evolve, we will likely see even more innovative applications of deep learning in healthcare.

46
 Deep Learning and Finance
• Deep learning has shown significant potential to revolutionize the financial industry, with a
wide range of applications that can help improve decision-making, risk management, fraud
detection, and customer experience. One of the key applications of deep learning in finance is
fraud detection. Deep learning models can be trained to identify fraudulent transactions,
reduce false positives, and improve accuracy. Deep learning algorithms can help financial
institutions to save millions of dollars in losses and improve customer trust and satisfaction.
• As mentioned, risk management is one of the applications of deep learning in finance. Deep
learning models can be used to analyze large datasets and identify patterns and correlations
that can help predict future risks and identify potential opportunities. This can help financial
institutions to make more informed decisions and reduce their risk exposure. Additionally,
deep learning can improve customer experience and satisfaction with boosting the accuracy
and efficiency of analyzing customer data and identifying behavior patterns.
• For example, deep learning models can be used to predict customer churn and provide
targeted offers and promotions to retain customers.
• Deep learning is poised to play an increasingly important role in finance, potentially
transforming how financial institutions operate and interact with customers.
47
 Deep Learning and Agriculture
• Deep learning has the potential to revolutionize the field of agriculture, enabling farmers to
make more informed decisions and increase crop yields. One of the key applications of deep
learning in agriculture is crop monitoring. Deep learning models can be trained to analyze
satellite or drone images and other types of remote sensing data to detect changes in crop
health and predict yields. This can be used to identify areas of the field that require
attention, such as where irrigation or fertilization is needed.
• Another important application of deep learning in agriculture is precision farming. Precision
agriculture uses data analytics and technology to optimize crop yield and reduce waste.
Deep learning models can be used to analyze data on factors such as soil moisture,
temperature, and nutrient levels. The models can then recommend when and where to plant,
irrigate, and fertilize crops. This can help farmers maximize resources and reduce costs,
improving their crops’ overall health and yield.
• Overall, deep learning can transform agriculture, helping farmers make more informed
decisions and achieve greater efficiency and sustainability in their operations.

48
 Deep Learning and Cybersecurity
• Cybersecurity also uses deep learning to improve threat detection and response. Cybersecurity
threats are becoming increasingly sophisticated, and traditional rule-based approaches are
insufficient to detect and prevent attacks. Deep learning algorithms can analyze large volumes of
data to identify patterns and anomalies that may indicate a security breach or attack.
• Intrusion detection is one application of deep learning in cybersecurity. Intrusion detection
systems (IDS) enable us to monitor network traffic and identify suspicious activity that may
indicate an attempted attack. Training deep learning algorithms on large network traffic datasets
for detecting patterns and anomalies that may indicate an attack allows us to identify and
respond to attacks more quickly and effectively than traditional rule-based approaches.
• Another application of deep learning in cybersecurity is in developing malware detection
systems. Malware is a type of malicious software that is used to compromise computer systems
and steal data. Deep learning algorithms can identify patterns that may indicate the presence of
malware, which leads to detecting and preventing malware attacks before they can cause
damage.
• Overall, deep learning is becoming an increasingly important tool in the field of cybersecurity,
enabling organizations to detect and respond to threats more quickly and effectively than ever
before. 49
 Self Driving Cars
• Deep Learning is the driving force behind the notion of self-driving
automobiles that are autonomous.
• Deep Learning technologies are actually "learning machines" that learn how
to act and respond using millions of data sets and training.
• To diversify its business infrastructure, Uber Artificial
Intelligence laboratories are powering additional autonomous cars and
developing self-driving cars for on-demand food delivery. Amazon, on the
other hand, has delivered their merchandise using drones in select areas of
the globe.
• The perplexing problem about self-driving vehicles that the bulk of its
designers are addressing is subjecting self-driving cars to a variety of
scenarios to assure safe driving. They have operational sensors for
calculating adjacent objects.
50
https://www.simplilearn.com/tutorials/deep-learning-tutorial/deep-learning-applications
https://www.slideshare.net/TundeAjoseIsmail/deep-learning-presentation-102934185 51
52
 Keras
• Keras is a high-level, deep learning API developed by Google for implementing
neural networks. It is written in Python and is used to make the implementation
of neural networks easy. It also supports multiple backend neural network
computation.
• Keras is relatively easy to learn and work with because it provides a python
frontend with a high level of abstraction while having the option of multiple back-
ends for computation purposes. This makes Keras slower than other deep
learning frameworks, but extremely beginner-friendly.
• Keras allows you to switch between different back ends. The frameworks
supported by Keras are:
• Tensorflow
• Theano
• PlaidML
• MXNet
• CNTK (Microsoft Cognitive Toolkit ) 53
Overview of Tools: Torch, TensorFlow, Keras

54
Deep Learning Neurons

55
https://towardsdatascience.com/whats-the-role-of-weights-and-bias-in-a-neural-network-4cf7e9888a0f
 Components of the basic Artificial Neuron:
• Inputs: Inputs are the set of values for which we need to predict a output
value. They can be viewed as features or attributes in a dataset.
• Weights: weights are the real values that are attached with each
input/feature and they convey the importance of that corresponding
feature in predicting the final output. (will discuss about this in-detail in
this article)
• Bias: Bias is used for shifting the activation function towards left or right,
you can compare this to y-intercept in the line equation.
• Summation Function: The work of the summation function is to bind the
weights and inputs together and calculate their sum.
• Activation Function: It is used to introduce non-linearity in the model.

56
Model of Artificial Neural Network

For the above general model of artificial neural network, the net input can be calculated as follows −

The output can be calculated by applying the activation function over the net input.

Output = function net input calculated 57

 The role of bias in Neural Networks
• The activation function in Neural Networks
takes an input 'x' multiplied by a weight 'w'.
Bias allows you to shift the activation
function by adding a constant (i.e. the given
bias) to the input. Bias in Neural Networks
can be thought of as analogous to the role of
a constant in a linear function, whereby the
line is effectively transposed by the constant
value.
• In a scenario with bias, the input to the
activation function is 'x' times the
connection weight 'w0' plus the bias times
the connection weight for the bias 'w1'.
This has the effect of shifting the
In a scenario with no bias, the input to the activation activation function by a constant amount
function is 'x' multiplied by the connection weight 'w0'.
(b * w1).
58
https://www.pico.net/kb/the-role-of-bias-in-neural-networks/
 Activation functions with Neural Networks
• It is used to determine the output of neural network like yes or no. It maps the
resulting values in between 0 to 1 or -1 to 1 etc. (depending upon the function).
• The Activation Functions can be basically divided into 2 types-
1. Linear Activation Function
2. Non-linear Activation Functions

https://towardsdatascience.com/activation-functions-neural-networks-1cbd9f8d91d6 59
https://www.researchgate.net/figure/Artificial-neural-network-activation-functions-In-this-figure-the-most-common_fig8_344331692