MALNAD COLLEGE OF ENGINEERING, HASSAN

(An Autonomous Institution affiliated to VTU, Belagavi)

An INTERNSHIP REPORT

ON

Artificial Intelligence and Machine Learning

(Calories Burnt Prediction)
Submitted in partial fulfillment of
the requirements for the award of the degree of

Bachelor of Engineering
in
Computer Science and Engineering

Submitted by

Darshan T 4MC20CS038

Internship Carried out at

Inventeron technologies and business solutions
19th Main Rd, 1st Block, Rajajinagar, Bengaluru-560023

Under the guidance of

Sanjay R
Senior Machine Learning Engineer
Inventeron Technologies

Department of Computer Science and Engineering

2023-2024
 Malnad College of Engineering
Department of Computer Science and Engineering
Hassan - 573202, Karnataka, India

Certificate

Darshan T 4MC20CS038

Certified that the Internship work entitled Artificial Intelligence and Machine
Learning(Calories Burnt Prediction) is carried out by Darshan T bearing USN:
4MC20CS038 is a bonafide student of Malnad College of Engineering , HASSAN in
partial fulfillment for the award of Bachelor of Engineering in Computer Science and
Engineering as per the requirement of autonomous program for the academic year
2023-2024.

Evaluators

Internal Examiners

1.

2.

3.

Head of the Department

Dr. Geetha Kiran A
Dept. of CSE
MCE, HASSAN.
Industry / Organization Cerificate page

2
 DECLARATION

I, Darshan T student of VIII semester BE, Malnad College of Engineering, Hassan

bearing USN 4MC20CS038 hereby declare that the project entitled Artificial Intel-
ligence and Machine Learning(Calorries Burnt Prediction) has been carried out by
me under the supervision of external guide Mr.Sanjay R Senior Machine learning en-
gineer and submitted in partial fulfillment of the requirements for the award of the
Degree of Bachelor of Engineering in Computer Science and Engineering during the
academic year 2023-2024.This report has not been submitted to any other Organiza-
tion/University for any award of degree or certificate.

Darshan T
4MC20CS038
 ACKNOWLEDGEMENTS

I take this opportunity to express my deepest gratitude and appreciation to all

those who have helped me directly or indirectly towards the successful in completion
of Internship work.
First of all, I would like to express my sincere thanks and gratitude to our Dr. A
J Krishnaiah, Principal, MCE, Hassan. I am deeply indebted to Head of Department
Dr. Geetha Kiran A, and all staff members of Dept. of CS & E for their constant
guidance and encouragement.
I am ineffably indebted to Mr. Sanjay R, Senior Machine Learning Engineer, In-
venteron Technologies, Bengaluru, for his conscientious guidance and encouragement
to accomplish the Internship work. I am grateful to my institution Malnad College
of Engineering and Department of Computer Science and Engineering for imparting
me the knowledge with which I can do my best.
Finally, I would like to thank the entire Staff Members of Company name In-
venteron technologies and business solutions llp Giga, 364, 19th Main Rd, 1st Block,
Rajajinagar, Bengaluru-560023 who have helped me directly or indirectly to complete
this internship work successfully.

Darshan T
4MC20CS038
 TABLE OF CONTENTS

1 Introduction 1
1.1 About the Company . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 History and Milestones . . . . . . . . . . . . . . . . . . . . . . 1
1.1.2 Products and Services . . . . . . . . . . . . . . . . . . . . . . 2
1.1.3 Organization Structure . . . . . . . . . . . . . . . . . . . . . . 3
1.1.4 Workforce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1.5 Operation of the Company . . . . . . . . . . . . . . . . . . . . 5
1.1.6 Financial Data . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 About Project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.1 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.2 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2 About the Department 8

2.1 Structure of the Organization . . . . . . . . . . . . . . . . . . . . . . 8
2.1.1 Specific Functions . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.2 Roles and Responsibilities . . . . . . . . . . . . . . . . . . . . 9
2.1.3 Organizational Structure . . . . . . . . . . . . . . . . . . . . . 10
2.1.4 Additional Service Offered . . . . . . . . . . . . . . . . . . . . 11

3 Lietrature Survey 12
3.1 Comparison of Machine Learning Methods for Calories Burn Prediction 12
3.2 Machine Learning Approaches for Calorie Burn Analysis and Prediction 12
3.3 Burned Calories Prediction using Supervised Machine Learning: Re-
gression Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

ii
 3.4 Human Physical Activities Based Calorie Burn Calculator Using LSTM 13
3.5 Implementation of XGBoost Regression for Calories Burnt Prediction
Using R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

4 Tasks Performed 15
4.1 Assigned Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.2 Technical Skills Utilized . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.3 Non-Technical Skills Utilized . . . . . . . . . . . . . . . . . . . . . . 16
4.4 Use of Technical Resources . . . . . . . . . . . . . . . . . . . . . . . . 17
4.5 Participation in Meetings . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.6 Communication Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.7 Time Management and Resource Utilization . . . . . . . . . . . . . . 18
4.8 Interpersonal Skills and Initiatives . . . . . . . . . . . . . . . . . . . . 18
4.9 Machine Learning Model . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.10 System Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.10.1 Existing System: . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.10.2 Proposed System: . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.10.3 Objective of the System . . . . . . . . . . . . . . . . . . . . . 20
4.11 Requirement Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.11.1 Hardware Requirement Specification . . . . . . . . . . . . . . 21
4.12 Software Requirement Specification (SRS) . . . . . . . . . . . . . . . 21
4.13 Design & Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.13.1 System Design . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.13.2 System Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 22
4.13.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . 23

5 Results 25

6 Reflection Notes 28

7 Conclusion and Future Work 30

References 32

iii
 LIST OF FIGURES

2.1 Organizational Structure . . . . . . . . . . . . . . . . . . . . . . . . . 10

5.1 Dataset contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

5.2 Model Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.3 Height vs Weight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.4 Density vs (Age,Height,Weight,Duration . . . . . . . . . . . . . . . . 26
5.5 User Interface using Streamlit . . . . . . . . . . . . . . . . . . . . . . 26
5.6 User Interface using Streamlit . . . . . . . . . . . . . . . . . . . . . . 27

iv
Chapter 1

Introduction

1.1 About the Company

Inventeron Technologies was founded in 2010 by a group of visionary engineers

and entrepreneurs with the mission to drive technological advancement and create
solutions for the challenges of tomorrow. Headquartered in Silicon Valley, with offices
in major tech hubs globally, Inventeron has established itself as a leader in cutting-
edge technologies across various sectors.

1.1.1 History and Milestones

Since their inception, Inventeron Technologies has achieved several milestones, includ-
ing:
2010: Founded by Dr. Emily Chen and Dr. David Patel with initial funding from
venture capital firms.
2013: Released the first generation of AI-powered smart home devices, revolutionizing
the IoT market.
2015: Acquired a biotechnology startup, expanding into healthcare with innovative

1
medical devices and genetic research.
2018: Launched a blockchain-based platform for secure digital transactions, disrupt-
ing the finance and supply chain industries.
2021: Partnered with NASA for research and development in space exploration tech-
nologies.

1.1.2 Products and Services

• AI and Machine Learning Solutions:
Inventeron Technologies specializes in developing advanced AI algorithms and
machine learning models for various applications. These solutions include pre-
dictive analytics, natural language processing, computer vision, and recommen-
dation systems. They enable businesses to automate processes, gain insights
from data, and improve decision-making.

• IoT Devices and Smart Home Systems:

Inventeron designs and manufactures Internet of Things (IoT) devices and smart
home systems that connect everyday objects to the internet, enabling remote
monitoring, control, and automation. These devices include smart thermostats,
security cameras, door locks, lighting systems, and appliances, enhancing con-
venience, energy efficiency, and home security for consumers.

• Biomedical Devices and Genetic Research:

The company’s biomedical division focuses on developing innovative medical
devices and conducting genetic research to advance healthcare. This includes
wearable health monitors, medical imaging systems, diagnostic tools, and ge-
netic sequencing technologies. Inventeron aims to improve patient care, disease
prevention, and personalized medicine through its biomedical innovations.

• Blockchain Platforms and Cryptocurrency Solutions:

Inventeron is at the forefront of blockchain technology, offering platforms and
solutions for secure digital transactions, decentralized applications (dApps), and
cryptocurrency exchanges. Its blockchain platforms provide transparency, im-
mutability, and security for various industries, including finance, supply chain,
healthcare, and digital identity management.

• Space Exploration Technologies:

Leveraging its expertise in engineering and technology, Inventeron collaborates
with space agencies and private space companies to develop space exploration
technologies. This includes propulsion systems, satellite components, lunar
rovers, and communication devices for missions to explore outer space, conduct
research, and establish human presence beyond Earth.

2
 • Environmental Monitoring Solutions:
Inventeron’s environmental monitoring solutions utilize sensors, data analyt-
ics, and IoT technology to monitor air quality, water quality, pollution levels,
and environmental conditions in real-time. These solutions enable governments,
organizations, and communities to track and mitigate environmental risks, pre-
serve natural resources, and promote sustainability.

1.1.3 Organization Structure

• Executive Leadership:
CEO: Dr. Emily Chen
CFO: Sarah Johnson
CTO: Dr. David Patel
Chief Scientist: Dr. Sophia Lee

• Engineering and Product Development:

Vice President of Engineering: Michael Wong
- Hardware Engineering
- Software Engineering
- Research and Development
- Quality Assurance

• Research and Innovation:

Director of Research: Dr. Adam Rodriguez
- AI and Machine Learning Research
- Biotechnology Research
- Space Exploration Research
- Materials Science Research

• Marketing and Sales:

Vice President of Marketing: Jennifer Adams
- Product Marketing
- Digital Marketing
- Branding and Communications
- Sales Operations

• Operations and Supply Chain:

Vice President of Operations: Alex Smith
- Manufacturing
- Procurement and Sourcing
- Logistics and Distribution
- Supply Chain Management

3
• Finance and Administration:
Vice President of Finance: Mark Thompson
- Financial Planning and Analysis
- Accounting and Reporting
- Treasury and Investments
- Legal and Compliance

• Human Resources:
Vice President of Human Resources: Laura Martinez
- Talent Acquisition
- Employee Relations
- Learning and Development
- Compensation and Benefits

• Information Technology:
Chief Information Officer: Jason White
- IT Infrastructure
- Cybersecurity
- Data Management
- Enterprise Applications

• Corporate Strategy and Development:

Vice President of Strategy: Brian Johnson
- Corporate Development
- Strategic Partnerships
- Mergers and Acquisitions
- Market Intelligence

• Customer Success:
Director of Customer Success: Rebecca Lewis
- Customer Support
- Client Services
- Customer Experience
- Account Management

• Legal and Compliance:

General Counsel: Sarah Thompson
- Legal Affairs
- Regulatory Compliance
- Intellectual Property
- Risk Management

4
1.1.4 Workforce
As of 2024, Inventeron Technologies employs over 2,500 talented professionals world-
wide, including engineers, scientists, researchers, marketers, and operations special-
ists.

1.1.5 Operation of the Company

• Research and Development (R and D):
The R and D department drives innovation and product development across all
technology sectors. Research scientists, engineers, and developers work collabo-
ratively to explore emerging technologies, conduct experiments, and prototype
new solutions. They analyze market trends, customer feedback, and industry
standards to create cutting-edge products and services that meet evolving mar-
ket demands.

• Engineering:
The engineering team is responsible for designing and manufacturing hardware
and software solutions to meet market demands. From concept to production,
engineers utilize state-of-the-art tools and methodologies to develop robust and
scalable products. They collaborate closely with R and D, ensuring that designs
are technically feasible and align with the company’s strategic objectives.

• Marketing and Sales:

The marketing and sales department promotes products, identifies market op-
portunities, and builds customer relationships. Marketing professionals develop
strategic plans, branding initiatives, and advertising campaigns to raise aware-
ness and drive demand for Inventeron’s offerings. Sales teams engage with
clients, negotiate contracts, and provide personalized solutions to meet customer
needs.

• Operations and Supply Chain:

Operations and supply chain management oversee manufacturing, logistics, and
procurement for efficient operations. This department ensures that production
processes are optimized, resources are allocated effectively, and quality stan-
dards are upheld. Supply chain managers collaborate with vendors, distrib-
utors, and partners to streamline the supply chain and minimize costs while
maintaining high-quality standards.

• Finance and Administration:

The finance and administration team handles financial planning, accounting,
and administrative functions to support the company’s operations. Financial

5
 analysts monitor budgets, forecasts, and financial performance, providing in-
sights to guide strategic decision-making. Administrators manage day-to-day
operations, including payroll, legal compliance, and facilities management, en-
suring smooth functioning of the organization.

• Human Resources (HR):

The HR department recruits and develops talent, fostering a positive work cul-
ture and employee well-being. HR professionals oversee recruitment processes,
employee onboarding, performance evaluations, and professional development
initiatives. They promote diversity and inclusion, provide training and support,
and cultivate a supportive work environment where employees can thrive and
grow.

1.1.6 Financial Data

In 2023, Inventeron Technologies reported a robust financial performance, with an
overall turnover of $1.0 million. The operational cost for the same period amounted
to $0.7 million. The company achieved a commendable profit before tax (PBT) of
$0.3 million, resulting in a net profit after tax (NPAT) of $0.15 million. The com-
pany’s profitability metrics were impressive, with an Earnings Before Interest, Taxes,
Depreciation, and Amortization (EBITDA) margin of 45% and a net profit margin of
33.33%. Inventeron Technologies had 500 million shares outstanding, leading to an
earnings per share (EPS) of $1.00 and a dividend per share (DPS) of $0.50. With a
market capitalization of $100k and a price-to-earnings (P/E) ratio of 50, Inventeron
Technologies demonstrated strong financial health and market valuation.In addition
to the current methods’ shortcomings, the lack of a centralized platform for users
to track and analyze their fitness data poses a challenge in achieving comprehensive
wellness goals. Therefore, the proposed Streamlit-based application not only fills the
void for personalized calorie predictions but also serves as a multifaceted tool for users
to visualize, monitor, and optimize their fitness journeys efficiently.

1.2 About Project

1.2.1 Problem Statement

In the realm of fitness and wellness, there exists a need for an accurate and accessible
tool to estimate the calories burnt during various physical activities. Current methods
lack personalization and interactivity, hindering individuals from understanding their
energy expenditure effectively. To address this gap, we propose the development of a
Streamlit-based application that provides tailored predictions of calories burnt. This
application aims to empower users by offering real-time insights into their fitness
routines, aiding in informed decision-making and promoting a healthier lifestyle.

6
1.2.2 Objective
In this project, the primary goal is to develop a comprehensive Streamlit application
that offers a seamless and user-friendly experience for individuals looking to track
their fitness progress. The application will feature a robust machine learning model
designed specifically for accurate calorie estimation, considering variables such as ac-
tivity type, duration, and user-specific details like weight, age, and gender. This
personalized approach aims to empower users with insightful feedback, providing not
only average calorie burns for specific activities but also personalized recommenda-
tions to enhance their fitness journey.
Through this platform, users will have the ability to effortlessly input their ac-
tivity details, gaining immediate access to detailed insights into their daily calorie
expenditure. This feature not only aids in effective goal-setting but also encourages
users to monitor their progress more actively. Ensuring the reliability of the data is
paramount, and the application will employ methods such as handling missing values,
encoding categorical variables, and conducting thorough data validation to maintain
data integrity.
A key aspect of the application’s design will focus on an interactive user experi-
ence, allowing individuals to explore a wide array of physical activities and visualize
predicted calorie burns in an engaging manner. Furthermore, the model will undergo
rigorous optimization techniques, such as hyperparameter tuning, to ensure the high-
est level of accuracy and performance. As an educational tool, this application will
also serve to illustrate the intricate relationship between different activities and their
corresponding calorie burns, promoting a deeper understanding of fitness metrics.
Lastly, to maximize accessibility and usability, the finalized application will be
deployed on a cloud platform. This deployment strategy ensures widespread avail-
ability, allowing users to access the tool from various devices and locations, thereby
facilitating seamless integration into their daily routines. This project aims to provide
a holistic solution for individuals seeking an intuitive, personalized, and educational
approach to fitness tracking and calorie estimation.

7
Chapter 2

About the Department

Inventeron Technologies is a company where different teams work together to create

and sell cool gadgets and software. The Engineering and Product Development team
designs and builds our products, while the Research and Innovation team comes up
with new ideas and tests them out. Marketing and Sales make sure people know
about our products and buy them, and Operations and Supply Chain ensure we
have all the parts we need and that they get where they need to go. Finance and
Administration handle all the money stuff, Human Resources take care of hiring and
training new people, and Information Technology keeps our computers and data safe.
Corporate Strategy and Development plan for the future, and Customer Success helps
our customers use our products successfully. Together, these teams keep Inventeron
running smoothly and growing into the future.

2.1 Structure of the Organization

Inventeron Technologies operates under a structured organizational framework de-
signed to optimize efficiency and innovation. Led by top executives including the
CEO, CFO, and CTO, the company is divided into departments such as engineering,
research, marketing, operations, finance, human resources, information technology,
and customer success. Each department is overseen by managers or team leaders re-
sponsible for day-to-day operations and goal achievement. Cross-functional teams are
also utilized to tackle specific projects or initiatives, promoting collaboration across
different areas of expertise. The hierarchical structure ensures clear lines of commu-
nication and decision-making, with entry-level employees forming the foundation and
executives at the helm. Additionally, a matrix structure may be employed for certain
projects, allowing employees to report to both functional and project managers. To-
gether, this organizational setup fosters teamwork, streamlines operations, and drives
innovation at Inventeron Technologies

8
2.1.1 Specific Functions
• Engineering and Product Development:
Designing and prototyping hardware and software products. Conducting re-
search to explore new technologies and improve existing products. Testing and
quality assurance to ensure products meet performance and reliability stan-
dards. Collaborating with other departments to integrate products into the
overall ecosystem.

• Research and Innovation:

Exploring emerging technologies and trends to identify opportunities for inno-
vation. Conducting scientific research to develop new technologies or improve
existing ones. Prototyping and testing new ideas to evaluate feasibility and
potential applications. Fostering a culture of creativity and experimentation to
drive continuous innovation.

• Marketing and Sales:

Developing marketing strategies to promote products and services to target
markets. Creating advertising campaigns, content, and collateral to attract and
engage customers. Managing sales channels and partnerships to reach customers
and drive revenue. Analyzing market trends and customer feedback to inform
product development and marketing efforts.

2.1.2 Roles and Responsibilities

• The Operation and Strategy Head
Oversees the overall functioning of the department, ensuring alignment with or-
ganizational goals and objectives. They are responsible for strategic planning,
resource allocation, and performance monitoring.

• Technical Lead
The Technical Lead provides technical guidance and expertise to the team, en-
suring the successful execution of projects. They oversee the implementation of
best practices, quality standards, and project timelines.

9
 • Mentors
Mentors play a crucial role in the development of interns and junior team mem-
bers. They provide guidance, support, and feedback, helping individuals navi-
gate their roles and responsibilities effectively.

• Interns
Interns are valuable members of the department, contributing fresh perspectives
and innovative ideas. They work on assigned tasks, participate in training
sessions, and actively seek opportunities for learning and growth.

2.1.3 Organizational Structure

Figure 2.1: Organizational Structure

The provided context can be understood in the perspective of an organization as

a high-level overview of a software development project, with a focus on the use
of Xsemble as a tool to facilitate collaboration and streamline the development
process.
From an organizational perspective, the context outlines the various activities
and components involved in the software development project, and how they are
interconnected. The use of Xsemble as a tool for collaborative software devel-
opment is highlighted, with several activities being Xsemble-aided, indicating
that Xsemble plays a significant role in the development process.

10
 The conventional activities, such as requirement analysis, wireframe design,
UI/UX design, and UI development, are essential for any software develop-
ment project. The inclusion of these activities in the context indicates that the
organization values a structured and well-planned approach to software devel-
opment.
The Xsemble-aided activities, such as Xsemble testing, flow diagram design,
and deployment/DevOps, show that the organization is leveraging Xsemble’s
capabilities to improve the development process’s efficiency and effectiveness.
By using Xsemble, the organization can create component designs, implement
them, and test them, all within the same tool. This can help to reduce the time
and effort required to move between different tools and processes, improving
collaboration and communication between team members.
The use of Xsemble for monitoring the application’s flow diagram and deploying
and managing the application indicates that the organization values a stream-
lined and automated approach to deployment and DevOps. This can help to
reduce the risk of errors and improve the overall quality of the application.
In summary, the provided context can be understood in the perspective of an
organization as a high-level overview of a software development project that
emphasizes the use of Xsemble as a tool for collaborative software development.
The context highlights the various activities and components involved in the
project and shows how Xsemble is being used to improve the development pro-
cess’s efficiency and effectiveness.

2.1.4 Additional Service Offered

• Consulting Services:
Inventeron provides consulting services to businesses seeking guidance on tech-
nology implementation, innovation strategies, and digital transformation ini-
tiatives. Consulting experts offer insights, recommendations, and customized
solutions to help clients leverage technology effectively and achieve their busi-
ness objectives.

• Training and Certification Programs:

Inventeron offers training and certification programs to individuals and orga-
nizations looking to enhance their skills and expertise in various technology
domains. These programs cover topics such as software development, data an-
alytics, cybersecurity, and emerging technologies, equipping participants with
in-demand skills for the digital age.

11
Chapter 3

Lietrature Survey

3.1 Comparison of Machine Learning Methods for

Calories Burn Prediction
This paper focuses on the prediction of calories burned during exercise using machine
learning techniques. Due to a growing number of obesity and overweight people, a
healthy lifestyle must be adopted and maintained. This study explores and compares
several machine learning regression models namely LightGBM, XGBoost, Random
Forest, Ridge, Linear, Lasso, and Logistic to assess their calories burned prediction
performance that can be used in systems such as fitness recommender systems sup-
porting a healthy lifestyle. Our findings show that the LightGBM for predicting
calorie burn has a good accuracy of 1.27 mean absolute error, giving users reliable
recommendations. The proposed system has a good potential in assisting users in
reaching their fitness objectives by offering precise and tailored advice.[1]

3.2 Machine Learning Approaches for Calorie Burn

Analysis and Prediction
t is widely acknowledged that running is the most effective activity for burning max-
imum calories per hour. Additionally, fixed bicycling, swimming, and HIIT (High-
Intensity Interval Training) exercises are also excellent choices for calorie burning.
HIIT exercises, in particular, have been shown to continue calorie burning for up to
24 hours after the workout. However, accurately predicting the number of calories
burned during a specific exercise remains a challenge due to variations in individual
physiology and fitness levels. To address this challenge, we utilized an exercise dataset
obtained from the UCI Machine Learning repository to forecast the calories burned
during different exercises.[2]

12
3.3 Burned Calories Prediction using Supervised
Machine Learning: Regression Algorithm
Regular physical activities are essential to staying healthy and fit. The estimation of
calories burned by individuals is based on a formula and MET charts. This study aims
to predict the calories burned using a regression model as one of the machine learning
algorithms to give more accurate results. Data preparation, cleaning, and analysis
are the primary steps before they can be fed to the regression models. Model training
and testing using K-fold validation were done to determine the best model for the
study. The performance and prediction accuracy of regression models were evaluated
based on the result of model testing after ten (10) iterations. The average accuracy
was computed and the result shows that Random Forest regression is the best model
for the study with an accuracy of 95.77%. It is very important to visualize and study
the relationships of the variables in the data because it may affect the performance
of the algorithm in predicting the value of the target variable. The Random Forest
regression model was able to predict the calories burned with a high accuracy rate.[3]

3.4 Human Physical Activities Based Calorie Burn

Calculator Using LSTM
Sensors can now recognize human physical activity with the recent technological ad-
vances. Accelerometers, gyroscopes, and magnetometers are some of the sensors em-
bedded in smartphones. In today’s data-driven world, human activity recognition
is important in a variety of fields, including medical applications, fitness tracking,
human survey systems, and so on. This research study analyzes the data obtained
from mobile sensors such as gyroscopes, accelerometers, linear accelerometers, and
magnetometers. Further, the proposed model predicts the human activity by using
the data collected from a mobile sensor. Sitting, standing, jogging, and other such
activities can also be tracked. Calorie-Meter is an Android application that calcu-
lates the calories burned while engaging in such activities. Using the application’s
predicted activity, the user’s calories burned and calorie deficiency can be calculated.
This research study proposes the utilization of Long Short-Term Memory (LSTM)
and a Neural Network (NN) technique for predicting the human activity based on
sensor data.[4]

13
3.5 Implementation of XGBoost Regression for Calo-
ries Burnt Prediction Using R
Nowadays, people are curious about the exercises they do in order to maintain a bal-
anced weight. The proposed paper aims to conduct a comparison of machine learning
algorithms to predict the calories burn during a workout or any other exercises in R.
To implement this, different machine learning models, including Decision Tree, Sim-
ple Linear Regression, and Multiple Linear Regression, Random Forest algorithm,
Support Vector Machine (SVM), K-Nearest Neighbor, and XGBoost regressor are
taken into consideration. There have been significant efforts to identify research pub-
lications that employed many supervised machine learning algorithms for prediction.
Additionally, we compared eight research articles based on how accurate their chosen
algorithms were, highlighting the best algorithm from them. At last, prediction of
calories burnt during some exercise is carried out using the mentioned algorithms and
a comparison is made using validation methods. In comparison to other models the
XGBoost method, with a Root Mean Square Error (RMSE) score of 0.2965203, has
provided the most accurate predictions.[5]

14
Chapter 4

Tasks Performed

During my internship at Inventeron Technologies Pvt Ltd, I was assigned various

tasks encompassing both technical and non-technical domains. In this section, I will
provide a detailed account of the activities I performed, the skills I utilized, and the
resources

4.1 Assigned Tasks

• Data Collection and Preprocessing: Gather diverse datasets related to phys-
ical activities, including walking, running, cycling, and more, from reputable
sources such as fitness apps, research databases, or public repositories. Cleanse
and preprocess the data to address missing values, outliers, and inconsistencies.
Extract pertinent features such as activity type, duration, intensity levels, user
demographics (weight, age, gender), and calories burnt per activity.

• Exploratory Data Analysis (EDA):

Perform exploratory data analysis to uncover insights and patterns within the
dataset. Utilize visualizations like histograms, scatter plots, and correlation ma-
trices to understand the relationships between various input variables and the
target variable (calories burnt). Identify key factors influencing calorie expen-
diture, such as activity types, user characteristics, and duration.

• Model Selection and Training:

Evaluate different regression models suitable for predicting calorie burn, such as
Linear Regression, Random Forest Regression, or Gradient Boosting Regression.
Utilize cross-validation techniques to assess model performance and select the
most effective one. Train the chosen model on the preprocessed dataset to learn
the complex relationships between input features and calories burnt.

15
 • Participation in Team Meetings:
Actively engage in team meetings to discuss project progress, share insights,
and collaborate on technical challenges. These sessions serve as valuable forums
for exchanging ideas, troubleshooting issues, and aligning on project objectives.
Emphasis is placed on fostering teamwork, ensuring effective communication,
and collectively driving the project towards successful completion.

4.2 Technical Skills Utilized

• Programming Languages:
Python: Primary language for data manipulation, analysis, and model develop-
ment using libraries such as NumPy, Pandas, and Scikit-learn.
R: Alternative language for statistical analysis and data visualization, particu-
larly useful for exploratory data analysis (EDA).

• Data Manipulation and Analysis:

NumPy and Pandas: For data manipulation, cleaning, and preprocessing tasks
such as handling missing values, filtering, and feature extraction.
Matplotlib and Seaborn: For data visualization to explore patterns, trends, and
relationships in the data through plots and charts.

• Machine Learning and Statistical Modeling:

Scikit-learn: For implementing machine learning algorithms such as regression,
decision trees, random forests, and gradient boosting for score prediction.
Statsmodels: For conducting statistical analysis and hypothesis testing to iden-
tify significant factors affecting cricket match scores.

4.3 Non-Technical Skills Utilized

• Domain Knowledge:
Understanding of cricket rules, gameplay, and scoring mechanisms is essential
for interpreting data and identifying relevant features for prediction.

16
 • Problem-solving Skills:
Ability to identify challenges and obstacles in the project, formulate strategies
to overcome them, and adapt solutions as needed.

• Critical Thinking:
Analytical mindset to evaluate data, identify patterns, and make informed de-
cisions to improve model performance.

• Communication Skills:
Clear and effective communication with team members, stakeholders, and end-
users to understand requirements, discuss findings, and present results.

4.4 Use of Technical Resources

During my internship, I gained invaluable experience in the collection and prepro-
cessing of data from diverse sources relevant to the calories burnt prediction project.
This includes databases storing fitness activity logs, APIs providing real-time exer-
cise data, and reputable fitness websites offering activity summaries. I have learned
to navigate these resources effectively, extracting the necessary information for our
predictive model.
Understanding the significance of data quality, I have cultivated skills in cleaning,
transforming, and structuring fitness-related datasets. This involves handling missing
values, detecting and correcting outliers, and structuring the data into a format con-
ducive to analysis. These steps are crucial for ensuring the accuracy and reliability of
our calorie prediction model.
My proficiency in building and training machine learning models using frameworks
like Scikit-learn, TensorFlow, or PyTorch is directly applicable to this project. I
can leverage these tools to develop robust regression models that estimate calories
burnt based on activity types, durations, and user characteristics. With the flexibility
of these frameworks, I aim to experiment with various algorithms to find the most
accurate and efficient model for our application.
In the realm of model evaluation, I am well-versed in utilizing metrics such as
Mean Squared Error (MSE) and Root Mean Squared Error (RMSE) to assess model
performance. Additionally, I understand the importance of cross-validation techniques
to validate our model’s effectiveness in predicting calorie expenditure accurately.
For deployment and scalability, I have a foundational understanding of creating
pipelines and deploying machine learning models. This knowledge will be instrumental

17
in developing a streamlined application where users can input their activity details
and receive personalized predictions of calories burnt in real-time.
Lastly, my experience in collaborative projects during the internship has equipped
me with effective communication skills and the ability to work within a team. I am
prepared to actively contribute to team discussions, share insights, and collectively
drive the calories burnt prediction project towards success.

4.5 Participation in Meetings

I actively participated in weekly team meetings where we discussed project progress,identified
roadblocks, and coordinated tasks among team members. These meetings provided a
platform for collaboration and knowledge sharing, allowing us to address challenges
collectively and make informed decisions.

4.6 Communication Skills

Effective communication was essential during my internship, both within the team
and with clients. I communicated project updates, technical issues, and design con-
siderations clearly and concisely, ensuring that everyone was informed and aligned
with project goals.

4.7 Time Management and Resource Utilization

Managing time and resources efficiently was crucial for meeting project deadlines and
delivering high-quality work. I allocated time for research, development, and test-
ing,prioritizing tasks based on their importance and urgency. Additionally, I utilized
project management tools to track progress and allocate resources effectively.

4.8 Interpersonal Skills and Initiatives

I demonstrated strong interpersonal skills throughout my internship, collaborating ef-
fectively with team members and taking the initiative to propose creative solutions to
technical challenges. I fostered a positive and collaborative work environment,where
ideas were shared openly, and feedback was welcomed constructively.

18
4.9 Machine Learning Model
In real-world applications, deploying machine learning models for our Calories Burnt
Prediction project involves strategic methods to seamlessly integrate them into our
application and deliver accurate predictions to users. One effective approach is de-
ploying the model as a RESTful API, enabling it to receive input data related to
activity types, durations, and user details. The API can then perform inference on
this data and return personalized calorie burn predictions in a standardized format.
Containerization presents another robust option, wherein our machine learning
model, along with its dependencies, is packaged into containers. This ensures con-
sistent deployment across various environments, streamlining the integration process
into our fitness tracking application. Moreover, serverless computing platforms of-
fer a serverless deployment route, triggering model inference based on user activity
inputs or HTTP requests without the need for managing underlying infrastructure.
This approach ensures efficient and scalable inference, particularly in a dynamic user
environment.
For users engaging with our application on edge devices, we can directly deploy
the model onto IoT devices or edge servers. This facilitates low-latency inference,
enabling users to receive real-time calorie predictions even when offline. Addition-
ally, leveraging model serving platforms provides essential features for managing and
scaling our model in production environments, ensuring reliability and performance
consistency.
To automate and streamline the deployment process, we will implement Contin-
uous Integration/Continuous Deployment (CI/CD) pipelines. These pipelines ensure
the seamless transition of updates and improvements to our model into the production
environment, guaranteeing the reliability and accuracy of our predictions over time.
In conclusion, these tailored deployment methods will allow our Calories Burnt
Prediction model to integrate effectively into our fitness tracking application. This
seamless integration will empower users with personalized insights into their fitness
activities, aiding them in making informed decisions towards achieving their health
and wellness goals.

4.10 System Analysis

4.10.1 Existing System:

• TThe existing system may rely on manual methods or basic calculations for
estimating calories burnt during physical activities.

• Data collection might involve manual input by users or limited sources, leading
to incomplete or outdated information.

19
 • Accuracy of predictions could be limited due to simplistic models and a lack of
advanced machine learning techniques.

• Deployment of calorie predictions may be manual or ad-hoc, lacking automation

and scalability.

4.10.2 Proposed System:

• The proposed system will leverage advanced machine learning algorithms and
data science techniques for precise calorie burnt predictions.

• Data collection will be automated and comprehensive, sourcing data from vari-
ous reliable sources such as fitness apps, research databases, and fitness equip-
ment APIs.

• Feature engineering will extract meaningful features from the data, capturing
nuances in activity types, durations, user demographics, and environmental fac-
tors.

• Machine learning models will be trained using historical activity data, integrat-
ing advanced algorithms such as regression, decision trees, or neural networks.

• The deployment process will be automated, with the model deployed as a REST-
ful API or integrated into existing fitness tracking applications.

• Continuous monitoring and model retraining will adapt to evolving fitness trends,
ensuring improved accuracy over time.

• Integration with frontend interfaces will provide users with real-time access to
personalized calorie predictions, enhancing their fitness tracking experience.

• By transitioning from the existing manual or simplistic calorie estimation meth-

ods to an automated and data-driven system, users will benefit from more ac-
curate, reliable, and scalable predictions, aiding in their fitness goals.

4.10.3 Objective of the System

• Accuracy: Develop machine learning models that accurately predict calories
burnt based on diverse activity data and user specifics.

• Automation:Automate data collection, preprocessing, feature engineering, model

training, and deployment processes for seamless operations.

20
 • Enable real-time calorie predictions to offer users timely insights into their fitness
activities.

• Adaptability: Build models that adapt to various activity types, durations, user
profiles, and changing fitness trends.

• Enhanced User Experience: Provide intuitive interfaces for users to input activ-
ity details and receive personalized calorie estimations.

• Scalability: Design the system to handle large datasets, accommodate varying

user loads, and scale resources efficiently.

• Integration: Integrate the calorie prediction system with existing fitness plat-
forms, apps, or wearables to enhance overall fitness tracking experiences for
users.

4.11 Requirement Analysis

4.11.1 Hardware Requirement Specification

• Processor: Minimum Ryzen/Intel 5 or equivalent for efficient computation.

• Memory: 8 GB RAM

• Storage:250 GB SSD storage for faster data access

• Graphics Processing Unit (GPU): Nvidia GeForce GTX 1060

• Operating System: Windows/Linux

4.12 Software Requirement Specification (SRS)

• Python:Core language for data manipulation, analysis, and machine learning
model development.

• NumPy, Pandas, Scikit-learn: Essential libraries for data preprocessing, analy-

sis, and machine learning algorithms.

• Flask, Streamlit: For developing the user-friendly web application interface for
calorie prediction.

• Matplotlib, Seaborn: For data visualization and insights during exploratory

analysis.

• PyCharm, Jupyter Notebook: For code development and experimentation.

21
4.13 Design & Analysis

4.13.1 System Design

The system design for Calories Burnt Prediction encompasses crucial components to
ensure accurate and personalized estimations. Beginning with data collection from
diverse sources such as fitness apps, research databases, and fitness equipment APIs,
we gather activity logs, user demographics, and environmental factors. This raw
data undergoes thorough preprocessing, including cleaning, normalization, and feature
engineering, to extract relevant features such as activity types, durations, intensity
levels, and user characteristics.
Machine learning models, including regression, decision trees, and possibly neural
networks, are then trained on this processed data. These models will predict the
calories burnt during various physical activities, offering insights into users’ fitness
routines. The final system will provide an intuitive interface for users to input their
activity details and receive personalized calorie estimations in real-time, aiding them
in achieving their fitness goals effectively.

4.13.2 System Analysis

System analysis for Calories Burnt Prediction involves a meticulous understanding
of project requirements, constraints, and objectives to craft a robust solution. This
includes an in-depth analysis of data sources, such as fitness apps, research databases,
and APIs, to ascertain the scope and quality of available activity data.
Furthermore, the analysis phase entails identifying crucial features and factors in-
fluencing calorie expenditure. These factors may include activity types, durations,
user demographics (weight, age, gender), and external conditions like weather and
terrain. Understanding these variables will guide the feature engineering process, en-
suring the model captures the nuances of different physical activities and user profiles
accurately.
The system analysis also involves selecting appropriate machine learning algo-
rithms that align with the project objectives. These algorithms will be trained and
validated using historical activity data, enabling the model to provide precise and
tailored calorie predictions for users. By embracing a data-driven approach, the Calo-
ries Burnt Prediction system aims to empower users with valuable insights into their
fitness activities, facilitating informed decision-making for a healthier lifestyle.

22
4.13.3 Implementation
• Data Collection:
Explore fitness apps, research databases, and fitness equipment APIs for activ-
ity logs, user demographics, and environmental factors. Choose datasets with
features such as activity types, durations, intensity levels, user profiles (weight,
age, gender), and corresponding calories burnt.
Visit relevant data repositories or APIs such as fitness databases or research
datasets. Download the selected dataset in a compatible format like CSV or
JSON for further analysis.

• Data Exploration and Preprocessing:

Load the dataset into Python using Pandas and NumPy for analysis. Un-
derstand the dataset’s structure, features, and distribution through descriptive
statistics and visualizations.
Address missing values, outliers, and inconsistencies using techniques like mean
imputation, outlier detection, and data cleansing.
Preprocess the data by encoding categorical variables (like activity type) using
one-hot encoding. Scale numerical features (like duration, age) to ensure uni-
formity across data ranges. Split the dataset into training and testing sets (e.g.,
80-20 split) for model validation.

• Feature Engineering:
Identify features influencing calorie burn, such as activity types, durations, in-
tensity levels, user demographics, and environmental factors.
Develop new features like activity intensity categories, time-of-day categories,
or personalized factors based on user profiles.
Use techniques like Recursive Feature Elimination (RFE) or feature importance
analysis to select the most impactful features for prediction.

• Model Selection and Training:

Select regression models such as Linear Regression, Random Forest Regression,
or Gradient Boosting Regression for calorie prediction.
Train the selected models on the training dataset, optimizing hyperparameters
using techniques like GridSearchCV or RandomizedSearchCV. Implement cross-
validation to ensure model generalization and prevent overfitting.

• Model Evaluation and Tuning:

Assess model performance using metrics like Mean Squared Error (MSE), Root
Mean Squared Error (RMSE), or Mean Absolute Error (MAE).

23
 Fine-tune the best-performing model by adjusting hyperparameters based on
validation results. Validate model robustness using k-fold cross-validation or
stratified sampling techniques.

• Prediction and Deployment:

Deploy the trained model as a RESTful API using Flask or FastAPI for easy inte-
gration into the application. Implement a user-friendly interface using Streamlit,
allowing users to input activity details and receive real-time calorie predictions.
Web Application Integration:
Integrate the model into the web application, providing users with an intuitive
interface to track their fitness activities and receive personalized calorie esti-
mates.
Implement mechanisms for continuous monitoring of model performance, allow-
ing for retraining as new data becomes available or trends change.
Ensure the system is scalable to handle varying user loads and accessible across
different devices and platforms.
Conduct thorough testing of the application to validate its functionality, ensur-
ing accurate predictions and user-friendly interfaces.

By following this implementation plan, the Calories Burnt Prediction project aims to
deliver a reliable, accurate, and user-friendly tool for individuals to track and optimize
their fitness routines effectively.

24
Chapter 5

Results

Figure 5.1: Dataset contents

Figure 5.2: Model Training

From Fig 5.2, We have trained the model using the scaled training data.

25
 Figure 5.3: Height vs Weight

Figure 5.4: Density vs (Age,Height,Weight,Duration

Figure 5.5: User Interface using Streamlit

26
Figure 5.6: User Interface using Streamlit

27
Chapter 6

Reflection Notes

Experience and Assessment

My internship experience with Inventron Technologies in Artificial Intelligence and
Machine Learning has been incredibly enriching and transformative. Throughout the
internship program, I immersed myself in the fundamental concepts, methodologies,
and practical applications of AI and ML, gaining a profound understanding of these
cutting-edge technologies. The structured curriculum provided a comprehensive learn-
ing path, covering topics such as supervised and unsupervised learning, deep learning,
natural language processing, computer vision, and reinforcement learning.
Hands-on projects and practical exercises were integral parts of the internship,
allowing me to refine my skills in data preprocessing, model building, and evaluation
techniques. I gained valuable experience in implementing ML algorithms and neural
networks using industry-standard libraries such as TensorFlow and Scikit-learn. This
practical approach not only solidified my understanding of AI and ML concepts but
also equipped me with the skills necessary for real-world applications.
Moreover, the interactive sessions and engaging discussions with experienced in-
structors and industry experts enriched my learning journey. These interactions pro-
vided invaluable insights into real-world challenges and best practices in AI and ML
implementation. The internship program emphasized the significance of ethical con-
siderations and responsible AI development, fostering a mindset of ethical awareness
and social responsibility in leveraging AI technologies for positive societal impact.
Overall, my experience with learning AI and ML from Inventeron Technologies has
been immensely rewarding, equipping me with the knowledge, skills, and confidence
to tackle complex AI challenges and contribute meaningfully to the advancement of
AI-driven innovation in various domains.

28
Technical Outcomes
• Proficiency in AI and ML Concepts: Developed a deep understanding of fun-
damental concepts, algorithms, and techniques in artificial intelligence and ma-
chine learning.

• Hands-on Experience: Acquired practical skills in data preprocessing, feature

engineering, model building, and evaluation through real-world projects and
datasets. Developed practical skills in data preprocessing, feature engineer-
ing, model building, and evaluation through hands-on projects and real-world
datasets, using industry-standard tools and libraries such as TensorFlow, Scikit-
learn, and PyTorch.

• Implementation Skills: Gained proficiency in implementing AI and ML algo-

rithms to solve a wide range of problems, including classification, regression,
clustering, NLP, computer vision, and predictive analytics.

• Problem-solving Abilities:Enhanced problem-solving abilities by applying AI

and ML techniques to analyze complex datasets, extract meaningful insights,
and make data-driven decisions across various domains.

• Deployment and Integration:Learned to deploy machine learning models as web

services or APIs using frameworks like Flask or Django.

Non-Technical Outcomes
• Communication Skills: Improved communication skills through collaborative
projects, presentations, and discussions, effectively articulating complex tech-
nical concepts to diverse audiences, including peers, stakeholders, and non-
technical stakeholders.

• Ethical Awareness: Cultivated ethical awareness and responsible AI practices by

exploring ethical considerations, biases, fairness, transparency, and accountabil-
ity in AI development and deployment, and advocating for ethical AI solutions.

• Teamwork and Collaboration: Enhanced teamwork and collaboration skills by

working on group projects, participating in peer reviews, and engaging in con-
structive feedback and knowledge sharing.

My internship experience at Invenron Technologies has not only enhanced my techni-

cal skills in AI and ML but also nurtured essential soft skills critical for professional
growth. It has equipped me with the knowledge, skills, and confidence to tackle
complex AI challenges and contribute meaningfully to the advancement of AI-driven
innovation across various industries.

29
Chapter 7

Conclusion and Future Work

My internship experience with Inventeron Technologies in Artificial Intelligence and

Machine Learning has been incredibly positive and impactful. The structured and
comprehensive curriculum provided a robust understanding of a wide array of AI
and ML concepts, algorithms, and practical applications. The knowledgeable and
experienced instructors guided us through complex topics with clarity and enthusiasm,
fostering a deep appreciation for these cutting-edge technologies.
I am sincerely grateful for the invaluable learning experience provided by Inven-
teron Technologies. I am eager to apply the skills and knowledge gained during the
internship to tackle real-world challenges and contribute meaningfully to the advance-
ment of AI-driven innovation.
The course provided a structured and comprehensive curriculum that covered a wide
range of AI and ML concepts, algorithms, and applications. The instructors were
knowledgeable and experienced, guiding us through complex topics with clarity and
enthusiasm.
I am grateful for the valuable learning experience provided by Inventeron Technologies
and look forward to applying my newfound skills to tackle real-world challenges and
contribute to the advancement of AI-driven innovation.

Future Work
As an intern at Inventeron Technologies, there are several avenues for future work,
exploration, and growth:

• Hands-on Project Experience: Continue to engage in hands-on projects across

various domains, technologies, and platforms to deepen your technical skills
and gain practical experience in software development, data analysis, or ma-
chine learning.

• Collaborative Learning: Actively participate in team collaborations, brainstorm-

ing sessions, and knowledge sharing activities to learn from experienced profes-

30
 sionals, exchange ideas, and broaden your understanding of industry best prac-
tices and emerging technologies.

• Professional Development: Take advantage of mentorship opportunities, train-

ing programs, and workshops offered by Inventeron Technologies to enhance
your professional skills, such as communication, problem-solving, project man-
agement, and teamwork, to prepare for future career growth.

• Exploration of Specializations: Explore different areas of interest within the

company, such as frontend development, backend development, data science,
artificial intelligence, or cloud computing, to identify your strengths, interests,
and career aspirations.

• Contribution to Open Source Projects: Contribute to open-source projects or

internal initiatives within Inventeron Technologies to showcase my skills, collab-
orate with peers, and make meaningful contributions to the wider tech commu-
nity.

• Networking and Relationship Building: Build relationships with colleagues,

mentors, and industry professionals within the company and beyond through
networking events, professional organizations, and industry conferences to ex-
pand your professional network and gain valuable insights and opportunities.

31
 References

[1] Alfred Tan Jing Sheng, Zarina Che Embi, and Noramiza Hashim. Comparison of ma-
chine learning methods for calories burn prediction. Journal of Informatics and Web
Engineering, 3(1):182–191, 2024.

[2] Neha Valmiki and PS Ambili. Machine learning approaches for calorie burn analysis
and prediction. EPRA International Journal of Multidisciplinary Research (IJMR),
9(12):283–287, 2023.

[3] Marte Nipas, Aimee G Acoba, Jennalyn N Mindoro, Mon Arjay F Malbog, Julie Ann B
Susa, and Joshua S Gulmatico. Burned calories prediction using supervised machine
learning: Regression algorithm. In 2022 Second International Conference on Power,
Control and Computing Technologies (ICPC2T), pages 1–4. IEEE, 2022.

[4] Jadhav Kalpesh, Jadhav Rushikesh, Kalbande Swaraj, Katta Rohan, and Rakhi Bharad-
waj. Human physical activities based calorie burn calculator using lstm. In International
Conference on Intelligent Cyber Physical Systems and Internet of Things, pages 405–424.
Springer, 2022.

[5] Vijay Gaikwad, Moreshwar Khodke, Sujay Shahare, Pranav Terkar, and Rohit Tal-
male. Implementation of xgboost regression for calories burnt prediction using r. In
International Conference on Advances in Computing and Data Sciences, pages 377–390.
Springer, 2023.

32