REPORT OF PRACTICAL TRAINING

on

ARTIFICIAL INTELLIGENCE & MACHINE LEARNING

Submitted In Partial Fulfilment of the requirements for the award of degree of
Bachelor of Technology
In
Computer Science & Engineering
From

Bikaner Technical University, Bikaner

(Session:2023-2027)

Submitted to:
Submitted by: Karan Solanki
Dr.Gagandeep Kaur
University Roll No: 23EBUCS033 Assistant Professor
Industrial Training Practical File Session 2024-25 Department of CSE
 DECLARTION

I, Karan Solanki, hereby declare that the Industrial Training report titled
“ARTIFICIAL INTELLIGENCE & MACHINE LEARNING” is an
authentic record of the work undertaken by me as part of the Industrial
Training. This report has been submitted in the partial fulfilment of the
requirements for the award of the Bachelor of Technology (Computer
Science and Engineering) degree. The information provided here is true
to the best of my knowledge.

Karan Solanki
Signature

Date:7-2-2025

ACKNOWLEDGEMENT

I convey my thanks to Dr. Gagandeep Kaur, Faculty of Computer

Science and Engineering, for providing me necessary support and details
at the right time during the progressive reviews.
S. N Table of Contents

➢ i Abstract

➢ ii Introduction of AI & ML

➢ iii Explain Block Chain Technology

➢ iv Differentiate between AR & VR

➢ v Define Cyber Security Attack and its Model.

➢ vi What is Phishing?

➢ vii Reference
 ABSTRACT
• This report presents an overview of the industrial training
undertaken at Bikaner Technical University focusing on
applications of artificial intelligence (AI) & machine
learning (ML) in practical environments. The primary
purpose of the training was to gain hands-on experience
in the field of AI & ML, with a specific focus on machine
learning, deep learning, and data analytics.

• The training provided invaluable insights into technical

aspects of AI & ML, development. Key takeaways included
the importance of data quality, the need for careful model
selection, and the role of evaluation metrics in assessing
model effectiveness. This experience has enhanced my
understanding of AI & ML, preparing me to contribute
effectively to future AI & ML projects.

INTRODUCTION OF AI & ML

1. Artificial intelligence (AI):

Artificial Intelligence (AI) refers to the simulation of human
intelligence processes by machines, especially computer systems.
AI systems are designed to perform tasks that typically require
human cognition, such as problem-solving, reasoning, learning,
decision-making, and natural language understanding. AI is broadly
classified into three categories:
1. Artificial Narrow Intelligence (ANI): AI specialized for specific
tasks, such as virtual assistants, recommendation systems, or
image recognition. This is the most common form of AI today.

2. Artificial General Intelligence (AGI): A theoretical form of AI

capable of performing any intellectual task that a human can do.

3. Artificial Super Intelligence (ASI): A level of AI surpassing human

intelligence in all aspects, including creativity, problem-solving,
and emotional intelligence.
➢ Key Components of AI:

• Data: Data serves as the foundation of AI. AI models learn

patterns and make predictions or decisions based on vast
amounts of structured or unstructured data.

• Algorithms: Algorithms process data and define the steps the AI

system must take to perform specific tasks.

• Models: AI models are the result of training algorithms on data.

These models are used to make predictions or automate
processes.

• Hardware and Infrastructure: Advanced computing power, GPUs,

TPUs, and cloud platforms enable the execution of complex AI
algorithms.

➢ Machine Learning (ML):

 Machine Learning (ML) is a subset of AI that enables systems
to learn and improve from data without being explicitly
programmed. ML algorithms find patterns and insights in data
and use them to make predictions or decisions. ML is integral
to AI, providing the learning mechanism that powers intelligent
systems.

• Types of Machine Learning:

1. Supervised Learning:

• The model is trained using labelled data, where inputs and

corresponding outputs are known.
• Example: Predicting house prices based on historical data (features:
location, size; output: price).

2. Unsupervised Learning:

• The model is trained on unlabelled data, identifying patterns or

clusters.
• Example: Customer segmentation in marketing based on purchasing
behaviour.

3. Reinforcement Learning:

• The model learns by interacting with an environment, receiving

rewards or penalties for actions.

• Example: Training an AI agent to play chess by optimizing moves for

victory.

Applications of machine learning in AI:

1. Natural Language Processing (NLP):

ML is used to analyse and understand human language. Applications

include chatbots, voice assistants, and sentiment analysis tools.

2. Computer Vision:

ML algorithms enable image recognition, object detection, and facial

recognition in fields like healthcare (diagnostic imaging) and security.

3. Recommendation Systems:

Platforms like Netflix and Amazon use ML to analyse user behaviour and
recommend relevant content or products.

4. Predictive Analytics:

ML models are used to forecast trends, such as stock prices, customer

churn, or disease outbreaks.

➢ Advantages of Using ML in AI:

Automation:
ML enables systems to perform repetitive tasks without manual
intervention.

Scalability:
AI systems powered by ML can handle vast amounts of data and scale
easily for large applications.
 Personalization:
ML allows AI to deliver personalized experiences by learning user
preferences.

Adaptability:
ML models adapt to new data, improving accuracy and relevance
over time.

Efficiency:
Automating processes with AI and ML reduces operational costs and
improves efficiency.
➢ Challenges of AI and ML:

1. Data Quality: Poor-quality data can lead to inaccurate

models and unreliable predictions.

2. Bias and Fairness: ML models can inherit biases present in the

training data, leading to unfair outcomes.

3. Interpretability: Understanding how complex ML models

make decisions (black- box models) remains challenging.

4. Ethical Concerns: The use of AI and ML raises questions about

privacy, job displacement, and misuse.

5. Resource Intensity: Training ML models requires significant

computational resources, which can be costly.

➢ Future of AI and ML:

The integration of AI and ML continues to transform industries and
redefine technological possibilities. Innovations such as Transformer
models (e.g., GPT, BERT) have revolutionized NLP, while
advancements in edge computing and federated learning promise
more secure and efficient AI applications. As AI evolves, the focus will
remain on creating systems that are not only intelligent but also
ethical, transparent, and beneficial to society.

➢ Training objective:

• The primary objective of my training at BTU was to gain

practical experience and foundational knowledge in artificial
intelligence (AI) & machine learning (ML). The training aimed
to bridge theoretical concepts with real-world applications,
equipping me with essential skills to handle data
preprocessing, model development, and evaluation.

• Another key objective was to develop problem-solving and

analytical skills essential for identifying and addressing AI &
ML-specific challenges, such as managing large datasets,
handling data inconsistencies, and optimizing model
performance. This aspect of the training was intended to
build a robust foundation for approaching complex AI & ML
problems systematically and efficiently.

• The training provided opportunities to delve deeply into

various ML models, such as logistic regression, decision
trees, support vector machines, and neural networks. A
particular focus was on NLP-specific architectures like
sequence-to-sequence models, attention mechanisms,
and Transformer-based models like BERT and GPT. These
models were used to improve key chatbot features like
intent recognition and context-aware response
generation.

• The training provided opportunities to delve deeply into various

ML models, such as logistic regression, decision trees, support
 vector machines, and neural networks. A particular focus was
on NLP-specific architectures like sequence-to-sequence
models, attention mechanisms, and Transformer- based models
like BERT and GPT. These models were used to improve key
chatbot features like intent recognition and context-aware
response generation.

• By the end of the training, my goal was to be proficient in

using AI tools, libraries to contribute effectively to
ongoing projects within a collaborative, professional
environment. This training provided me with the
technical and collaborative skills necessary to pursue
future AI-related roles with confidence and competence.

➢ BLOCK CHAIN TECHNOLOGY:

• Blockchain is a method of recording information that makes it

impossible or difficult for the system to be changed, hacked, or
manipulated. A blockchain is a distributed ledger that duplicates
and distributes transactions across the network of computers
participating in the blockchain.

• Blockchain technology is a structure that stores transactional

records, also known as the block, of the public in several
databases, known as the “chain,” in a network connected through
peer-to-peer nodes. Typically, this storage is referred to as a digital
ledger.
Every transaction in this ledger is authorized by the digital
signature of the owner, which authenticates the transaction and
safeguards it from tampering. Hence, the information the digital
ledger contains is highly secure.
• In simpler words, the digital ledger is like a Google spreadsheet
shared among numerous computers in a network, in which, the
transactional records are stored based on actual purchases. The
fascinating angle is that anybody can see the data, but they can’t
corrupt it.

❖ WORKING PROCESS OF BLOCK – CHAIN TECHNOLOGY

One of the famous uses of Blockchain is Bitcoin. Bitcoin is a

cryptocurrency and is used to exchange digital assets online.
Bitcoin uses cryptographic proof instead of third-party trust for
two parties to execute transactions over the Internet. Each
transaction protects through a digital signature.
 ❖ BENEFITS OF BLOCKCHAIN

• Time-saving: No central Authority verification is needed for settlements

making the process faster and cheaper.
• Cost-saving: A Blockchain network reduces expenses in several ways. No
need for third-party verification. Participants can share assets directly.
Intermediaries are reduced. Transaction efforts are minimized as every
participant has a copy of the shared ledger.
• Tighter security: No one can tamper with Blockchain Data as it is shared
among millions of Participants. The system is safe against cybercrimes and
Fraud.
• Collaboration: It permits every party to interact directly with one another
while not requiring third-party negotiation.
• Reliability: Blockchain certifies and verifies the identities of every
interested party. This removes double records, reducing rates and
accelerating transactions.

❖FUTURE SCOPE OF BLOCKCHAIN

TECHNOLOGY

• Digital Identity: Blockchain-based digital IDs might be used to store

personal data safely and securely as well as offer a means of
establishing identity without the need for a central authority.

• Smart Contracts: A variety of legal and financial transactions could

be automated using smart contracts, self-executing contracts with
the terms of the agreement put straight into lines of code.

• Decentralized Finance (DeFi): Using blockchain technology,

decentralized financial systems might be built that support peer-to-
peer transactions and do away with conventional intermediaries like
banks.

• Supply Chain Management: Blockchain technology can be applied

to a permanent record of how goods and services have been moved,
enabling improved openness and traceability across the whole
supply chain.
 Internet of Things (IoT): Blockchain technology may be used
to build decentralized, secure networks for IoT devices, enabling
them to exchange data and communicate with one another in an
anonymous, safe manner.

❖ Advantages of Blockchain Technology:

1. Decentralization: The decentralized nature of blockchain technology

eliminates the need for intermediaries, reducing costs and
increasing transparency.
2. Security: Transactions on a blockchain are secured through
cryptography, making them virtually immune to hacking and fraud.
3. Transparency: Blockchain technology allows all parties in a
transaction to have access to the same information, increasing
transparency and reducing the potential for disputes.
4. Efficiency: Transactions on a blockchain can be processed quickly
and efficiently, reducing the time and cost associated with
traditional transactions.
5. Trust: The transparent and secure nature of blockchain technology
can help to build trust between parties in a transaction.

❖ Disadvantages of Blockchain Technology:

1. Scalability: The decentralized nature of blockchain technology can

make it difficult to scale for large-scale applications.
2. Energy Consumption: The process of mining blockchain transactions
requires significant amounts of computing power, which can lead to
high energy consumption and environmental concerns.
3. Adoption: While the potential applications of blockchain technology
are vast, adoption has been slow due to the technical complexity
and lack of understanding of the technology.
4. Regulation: The regulatory framework around blockchain
technology is still in its early stages, which can create uncertainty for
businesses and investors.
 5. Lack of Standards: The lack of standardized protocols and
technologies can make it difficult for businesses to integrate
blockchain technology into their existing systems.
6. Overall, the advantages of blockchain technology are significant and
have the potential to revolutionize many industries. However, there
are also several challenges and disadvantages that must be
addressed before the technology can reach its full potential.

➢ INTRODUCTION OF AR & VR

AR→ Almost any person with a smartphone can get access

to augmented reality, making it more efficient than VR as
a branding and gaming tool. AR morphs the mundane,
physical world into a colourful, visual one by projecting
virtual pictures and characters through a phone's camera
or video viewer. Augmented reality is merely adding to
the user's real-life experience.

VR→ Virtual reality takes these same components to

another level by producing an entirely computer-
generated simulation of an alternate world. These
immersive simulations can create almost any visual or
place imaginable for the player using special equipment
such as computers, sensors, headsets, and gloves.

❖Difference Between AR and VR:

Virtual Reality (VR) Augmented Reality (AR)
Virtual Reality creates a fully Augmented Reality overlays digital
immersive digital environment or information into the real world.
experience that simulates the real
world or imaginary world.

It generally requires a headset or a It can be accomplished through

similar kind of device to immerse the smartphones or tablets with the help of AR
user into the digital world. apps.
The user is isolated from the real The user is aware of the real world while
world while in VR. experiencing AR.
It requires powerful hardware and It requires relatively simple technology for
software to create a realistic the creation.
experience.
Examples: PlayStation VR, Samsung Examples: Pokemon GO, Google Maps AR,
Gear VR, and HTC Vive. and IKEA App.

According to Tim Cook, CEO of Apple Inc.

“I think that a significant portion of the population of developed
countries, and eventually all countries, will have AR experiences
every day, almost like eating three meals a day. It will become that
much a part of you.”

❖Advantages of Augmented Reality

and Virtual Reality
 The VR and AR industry presents endless opportunities in
various fields. Virtual reality and augmented reality
enhance people's experience in apps, games, and movies
in the entertainment industry can help students know the
human body better in the medical field, and doctors can
train and practice dangerous surgeries without risk. VR is
also useful in schools, military training and space
programs. Virtual reality allows scientists to perform
experiments in an artificial environment. The business
sector can use augmented reality to develop product
designs, and the construction industry can use
augmented reality to identify and fix architectural flaws
of a building.

➢ CYBER ATTACK AND ITS MODEL

❖ INDRODUCTION: - A cyberattack – also known as a cybersecurity

attack – is any form of malicious activity targeting IT systems and/or the
attackers or threat actors using them to gain unauthorized access to
systems and data they contain.
Criminals typically are looking to exploit an attack for financial gain, but
in other cases the aim is to disrupt operations by disabling access to IT
systems. Threat actors can be anyone from a single person attempting to
obtain stolen credentials and hold them for ransom to a state-
sponsored contingent looking to disrupt operations on foreign soil.

❖COOMAN CYBER ATTACK MODELS:

When a criminal is trying to hack an organization, they won't try

something novel unless necessary. They draw upon common hacking
techniques that are known to be highly effective, such as malware or
phishing.
Whether you're trying to make sense of the latest data-breach headline
in the news or analysing an incident in your own organization, it helps
to understand different cyberattack vectors.

1. Malware:
Malware refers to various forms of harmful software, such as viruses
and ransomware. Once it is in your computer, it can wreak all sorts of
havoc, from taking control of your machine, to monitoring your actions
and keystrokes, to silently sending all sorts of confidential data from
your computer or network to the attacker's home base.
Attackers will use a variety of methods to get malware into your
computer, but at some stage it often requires the user to take an action
to install the malware. This can include clicking a link to download a file
or opening an email attachment that may look harmless (like a
document or PDF) but contains a hidden malware installer.

2. Ransomware:
Ransomware is a form of malware that encrypts data on infected IT
systems. It demands a ransom in exchange for a code that will –
hopefully – decrypt the infected system. The ransom payment usually
goes to an anonymous address using Bitcoin.

3. Adware:
Adware is a type of malware that displays unwanted ads on end-user
devices to generate revenue from advertisers. It often will be installed
on user devices after tricking people into clicking a link. Adware then
displays the ads and simulates user clicks to defraud advertisers into
thinking that legitimate users are interacting with their ads. They then
pay the cybercriminals for these clicks.

4. Crypto-Jacking:
Crypto jacking is a type of malware that uses the resources of the
infected IT systems to “mine” for cryptocurrencies. This steals the
attacked system's computing resources by running at a high load to
generate income for the remote attackers. They’ll then make money
from the sale of the cryptocurrencies generated on the infected
system.
5. Phishing:
In a phishing attack, an attacker may send you an email that appears to
be from someone you trust, like your boss or a company you do
business with. The email will seem legitimate, and it will have some
urgency to it (e.g. fraudulent activity has been detected on your
account). In the email, there may be an attachment to open or a link to
click.
Upon opening the malicious attachment, you'll unknowingly install
malware in your computer. If you click the link, it may send you to a
legitimate-looking website that asks you to log in to access an
important file – except the website is a trap used to capture your
credentials.

6. Spear Phishing:
Spear Phishing is a highly targeted variant of phishing that uses a fake
email or message from a supposedly important individual to trick a
person within the same organization or a partner organization. Spear
phishing attempts hope to use the extra authenticity – albeit imposter
authenticity – of the sender to trick people into providing information
they shouldn't.

7. SQL Injection Attack:

A structured query language (SQL) injection attack specifically targets
servers storing critical website and service data. It uses malicious code
to get the server to divulge information it normally wouldn’t. SQL is a
programming language used to communicate with databases and can
be used to store private customer information such as credit card
numbers, usernames and passwords (credentials), or other personally
identifiable information (PII) – all tempting and lucrative targets for an
attacker.

8. Cross-Site Scripting (XSS):

Cross-site Scripting attack also involve injecting malicious code into a
website, but in this case the website itself is not being attacked.
Instead, the malicious code only runs in the user's browser when they
visit the attacked website, where it directly targets the visitor.
One of the most common ways an attacker can deploy an XSS attack is
by injecting malicious code into a comment or a script that could
automatically run.
9. Botnets:
Botnets are widespread groups of devices that have been
compromised and hijacked by cybercriminals. The threat actors use
them to target IT systems with distributed DoS attacks or other attack
types.

10.Denial-of-Service (DoS):
Denial-of-Service (DoS) flood a website with more traffic than it’s built
to handle, thereby overloading the site’s server and making it near-
impossible to serve content to visitors. It’s possible for a denial-of-
service to occur for non-malicious reasons. For example, if a massive
news story breaks and a news organization’s site is overloaded with
traffic from people trying to learn more about the story.

11.Man In the Middle Attack:

A man in the middle attack (MITM) occurs when cybercriminals
intercept and alter network traffic flowing between IT systems. The
MITM attack impersonates both senders and receivers on the network.
It aims to trick both into sending unencrypted data that the attacker
intercepts and can use for further attacks or financial gain.

12.Session Hijacking:
Session hijacking occurs when an attacker hijacks a session by
capturing the unique – and private – session ID and poses as the
computer making a request, allowing them to log in as an unsuspecting
user and gain access to unauthorized information on the web server. If
everything goes as it should during any internet session, web servers
should respond to your various requests by giving you the information
you're attempting to access.

13.Credential Reuse:
Credential reuse occurs when someone uses the same credentials on
multiple websites. It can make life easier in the moment but can come
back to haunt that user later. Even though security best practices
universally recommend unique passwords for all applications and
websites, many people still reuse their passwords. This is a fact
attackers will readily exploit, thereby turning those reused passwords
into compromised credentials.
14.Insider Threats:
Not all cyber threats originate from external sources. Data and other
sensitive information like login credentials can leak from inside
organizations. This can occur via malicious staff activity or – more
frequently – due to an unintended action. An example of such a
mistake could be sending an email containing an unencrypted
attachment to the wrong recipient.

➢ PHISHING:

Phishing is another type of cyber-attack. Phishing got its name from

“phish” meaning fish. It’s a common phenomenon to put bait for the
fish to get trapped. Similarly, phishing works. It is an unethical way to
dupe the user or victim to click on harmful sites. The attacker crafts the
harmful site in such a way that the victim feels it to be an authentic
site, thus falling prey to it. The most common mode of phishing is by
sending spam emails that appear to be authentic and thus, taking away
all credentials from the victim. The main motive of the attacker behind
phishing is to gain confidential information like:

• Password
• Credit card details
• Social security numbers
• Date of birth

The attacker uses this information to further target the user

impersonate the user and cause data theft. The most common type of
phishing attack happens through email. Phishing victims are tricked
into revealing information that they think should be kept private. The
original logo of the email is used to make the user believe that it is
indeed the original email. But if we carefully investigate the details, we
will find that the URL or web address is not authentic. Let’s understand
this concept with the help of an example:
In this example, most people believe it’s YouTube just by looking at the
red icon. So, thinking of YouTube as a secure platform, the users click
on the extension without being suspicious about it. But if we look
carefully, we can see the URL is supertube.com and not youtube.com.
Secondly, YouTube never asks to add extensions for watching any
video. The third thing is the extension name itself is weird enough to
raise doubt about its credibility.
 ➢ REFERENCE:

Below is a list of references used during the AI training and chatbot project
development. These references include books, online documentation, and
tutorials that provided foundational knowledge, methodologies, and tools
relevant to artificial intelligence and natural language processing:

1. Jurafsky, & Martin (2020). Speech and Language Processing: An Introduction

to Natural Language Processing, Computational Linguistics, and Speech
Recognition. 3rd Edition. Pearson.
• This textbook was instrumental in understanding the theoretical
foundations of natural language processing (NLP) and its application to
tasks like sentiment analysis and intent recognition.

2. Harris (1954). Distributional Structure. Word.

• A seminal paper on distributional semantics that helped build the basis for
understanding how words are represented in high-dimensional spaces,
which is crucial for techniques like Word2Vec and TF-IDF.

3. Chollet (2017). Deep Learning with Python. Manning Publications.

• This book provided a hands-on approach to implementing deep learning
models, with practical examples using Keras and TensorFlow, which were
critical in training the intent classification models for the chatbot.