A PROJECT REPORT ON

HEART DISEASE PREDICTION USING K-NEAREST

NEIGHBORS (KNN) ALGORITHM
SUBMITTED TO THE SAVITRIBAI PHULE PUNE UNIVERSITY, PUNE
IN THE PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE AWARD OF THE DEGREE

BACHELOR OF ENGINEERING (ARTIFICIAL

INTELLIGENCE & DATA SCIENCE)

SUBMITTED BY
SAKSHI DIPAK SHINDE

DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND

DATA SCIENCE
PARVATIBAI GENBA MOZE COLLEGE OF ENGINEERING

BAIF ROAD, WAGHOLI, PUNE 412207

SAVITRIBAI PHULE PUNE UNIVERSITY

2024 -2025

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

1
 CERTIFICATE

This is to certify that the report entitles

“HEART DISEASE PREDICTION USING K-NEAREST NEIGHBORS (KNN)

ALGORITHM”

Submitted by

SAKSHI DIPAK SHINDE

is a bonafide student of this institute and the work has been carried out by her under the
supervision of Prof. Jaishri Shilpakar and it is approved for the partial fulfillment of the
requirement of Savitribai Phule Pune University, for the award of the degree of Bachelor of
Engineering (Artificial Intelligence and Data Science).

[Link] Shilpakar [Link] Shilpakar

Guide Head,
Department of Computer Engineering Department of Computer Engineering

Principal

Place : Pune
Date :

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

2
 TABLE OF CONTENTS

CHAPTER TITLE PAGE NO.

[Link]. Title of Chapter Page No.

01 Introduction 6
02 Title/Problem statement/objectives 8
03 Motivation/Scope and rationale of the study 9
04 Methodological details 11
05 Results / Analysis /inferences and conclusion 18
Acknowledgement 19
List of reference (Library books, magazines and other sources 20

LIST OF ABBREVIATION

ABBREVIATION ILLUSTRATION
ML Machine Learning
AI Artificial Intelligence
CSV Comma-Separated Value
KNN K-Nearest Neighbors

LIST OF FIGURES
FIGURE ILLUSTRATION PAGE NO.

1.1 System Overview 6

1.2 System Behavior 7
4.2 Timeline Chart 13
4.3 DFD Level – 0 14
4.4 DFD Level – 1 14
4.5 DFD Level – 2 15
4.6 Use case Diagram 15
4.7 Sequence Diagram 16
4.8 ER Diagram 16
4.9 Class Diagram 17

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

3
 ACKNOWLEDGEMENT

The success of any project depends largely on the encouragement and guidance of many
individuals. This research project would not have been possible without their support. I take
this opportunity to express my gratitude to those who have been instrumental in the successful
completion of this project. First and foremost, I wish to record my sincere gratitude to the
Management of this college and to my respected Principal, for his constant support and
encouragement in the preparation of this report and for providing the necessary facilities that
contributed to the successful completion of this project.
My sincere thanks to [Link] Shilpakar, Head, Artificial Intelligence and Data Science
, Parvatibai Genba Moze College Of Engineering, who was abundantly helpful and offered
invaluable assistance, support, and guidance throughout this research.
I am deeply grateful to my project guide, for her valuable insights, constant encouragement,
and continuous guidance throughout the investigation and implementation of this project on
Heart Disease Prediction using KNN Algorithm. Her expertise and supervision played a
crucial role in the successful execution of this work.
Lastly, I extend my heartfelt thanks to my faculty members, friends, and family for their
unwavering support, motivation, and encouragement, which have been a source of strength
throughout this journey.

NAME OF THE STUDENT:-

Shinde Sakshi Dipak

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

4
 ABSTRACT

Heart disease remains one of the leading causes of mortality worldwide, emphasizing the need
for early and accurate detection to improve patient outcomes. Traditional diagnostic methods
often rely on manual evaluation, which can be time-consuming and prone to human error. To
address these challenges, machine learning techniques are increasingly being adopted for
predictive analysis in healthcare.
This project presents a Heart Disease Prediction System using the K-Nearest Neighbors (KNN)
algorithm, which classifies patients based on critical medical parameters such as age, blood
pressure, cholesterol levels, and other cardiovascular risk factors. The model analyzes patient
data, identifies patterns, and predicts the likelihood of heart disease based on its proximity to
similar cases in the dataset.
The system's performance is evaluated using key classification metrics, including accuracy,
precision, recall, and F1-score, ensuring its reliability as a predictive tool. The results
demonstrate that the KNN-based approach provides a cost-effective and efficient solution for
early heart disease detection, aiding healthcare professionals in decision-making and
contributing to better disease management.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

5
 [Link].

1.1 SYSTEM BEHAVIOR

The behavior of the system describes how it interacts with users, processes data, and produces
predictions. The system follows a structured workflow to ensure accuracy, efficiency, and
usability.
1.1.1 Input Stage
• The system receives patient health data such as blood pressure, cholesterol, heart rate,
and other key indicators.
• The data can be manually entered by users (e.g., doctors, technicians) or fetched from
an electronic medical record system.
1.1.2. Data Processing & Feature Selection
• The system checks for missing values and applies data normalization to ensure
consistency.
• Key features affecting heart disease risk are selected for analysis.
1.1.3. Prediction Using KNN Algorithm
• The system computes the Euclidean distance between the new patient data and existing
medical records.
• It identifies the K nearest neighbors and assigns a classification based on the majority
class.
• The system then determines whether the patient is at High Risk, Low Risk, or No Risk
of heart disease.
1.1.4. Output & Decision Support
• The final prediction is displayed along with a confidence score.
• The system may recommend further medical tests or lifestyle modifications based on
the results.
• Doctors can use the prediction to assist in making early intervention decisions.
1.1.5. Error Handling & User Feedback
• If incomplete or incorrect data is entered, the system prompts the user to correct the
input.
• The system can adapt to new patient records by updating its dataset over time.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

6
1.2SYSTEM OVERVIEW.

The Heart Disease Prediction System using the K-Nearest Neighbors (KNN) algorithm is
designed to assist in the early diagnosis of heart disease by analyzing key medical parameters.
Traditional diagnostic methods can be time-consuming and dependent on expert evaluation,
leading to possible delays in treatment. This system leverages machine learning techniques to
classify patients based on their risk of heart disease, improving early detection and medical
decision-making.
1.2.1 Key Features:
• User-Friendly Input Interface: Allows users to enter medical parameters such as age,
blood pressure, cholesterol level, and heart rate.
• Efficient Data Processing: Cleans, normalizes, and processes input data for better
accuracy.
• Machine Learning-Based Prediction: Uses the KNN algorithm to classify patients based
on historical data.
• Real-Time Risk Assessment: Provides immediate prediction of heart disease likelihood.
• Decision Support System: Assists healthcare professionals in making informed
decisions.
1.2.2 System Components:
• Input Module: Accepts patient details such as age, cholesterol level, and other medical
attributes.
• Preprocessing Module: Handles missing values, normalizes data, and selects relevant
features.
• Prediction Model: Applies the KNN classification algorithm to determine the risk level.
• Output Module: Displays prediction results (e.g., High Risk, Low Risk, No Risk) and
confidence scores.
• Error Handling & Adaptability: Ensures data validation and updates the model with
new patient records.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

7
 [Link]

Heart Disease Prediction Using K-Nearest Neighbors (KNN) Algorithm

2.1 PROBLEM STATEMENT

Heart disease is a leading cause of death worldwide, and early detection is crucial for improving
patient outcomes. Traditional diagnostic methods require expert analysis, advanced medical
tests, and significant time, making early intervention challenging.

This project focuses on developing a machine learning-based prediction model using the K-
Nearest Neighbors (KNN) algorithm to analyze patient health parameters and predict the
likelihood of heart disease. The proposed system will provide an efficient and accessible
decision-support tool for healthcare professionals, assisting in early diagnosis and reducing
dependency on expensive medical tests.

2.2 OBJECTIVES

The main objectives of this project are:

• To develop a machine learning model using the K-Nearest Neighbors (KNN) algorithm
for heart disease prediction.

• To analyze key patient health parameters such as age, blood pressure, cholesterol levels,
heart rate, etc. for classification.

• To improve the accuracy of predictions by selecting an optimal K-value and refining

feature selection.

• To design a user-friendly interface for easy data input and prediction interpretation.

• To compare the performance of KNN with other classification algorithms and assess
its effectiveness.

• To assist in early detection and proactive healthcare decision-making using machine

learning.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

8
 3: MOTIVATION, SCOPE, AND RATIONALE OF THE STUDY

3.1 MOTIVATION
Heart disease remains one of the leading causes of death worldwide, with millions of lives lost
each year due to late diagnosis and lack of timely intervention. Traditional diagnostic methods
often require expensive medical tests, specialist consultations, and significant time for analysis,
making early detection challenging.

With the rise of machine learning and artificial intelligence, healthcare has the potential to
become more efficient, data-driven, and predictive. This project is motivated by the need to
develop a cost-effective, accessible, and accurate method for early detection of heart disease,
allowing individuals and healthcare professionals to take preventive measures before the
disease reaches a critical stage.

3.2 SCOPE OF THE STUDY

This study focuses on developing a predictive model for heart disease diagnosis using the KNN
algorithm. The dataset used for this project includes patient health parameters such as blood
pressure, cholesterol levels, heart rate, and other relevant attributes. The project involves:

• Data preprocessing to clean and prepare the dataset for training.

• Feature selection to identify the most relevant health parameters for prediction.

• Implementation of the KNN algorithm to classify patients as having heart disease or

not.

• Performance evaluation using accuracy, precision, recall, and F1-score.

• Development of a user-friendly interface (if applicable) for easy access to predictions.

• Comparison of KNN with other classification models to analyze its efficiency.

The ultimate goal is to enhance decision-making in healthcare by providing a reliable and

efficient tool for early detection of heart disease.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

9
3.3.3 RATIONALE OF THE STUDY
The rationale behind this study is to explore the practical application of machine learning in
healthcare, specifically in the early detection of heart disease. Given the growing availability
of health data and the advancements in AI, machine learning models can significantly enhance
decision-making, reduce manual errors, and assist healthcare professionals in identifying high-
risk patients earlier.

This project aims to:

1. Reduce the burden on healthcare professionals by providing an automated decision-

support system.

2. Improve early detection to help patients take preventive measures.

3. Enhance accessibility by offering a low-cost, data-driven prediction system.

4. Encourage future research in AI-driven healthcare diagnostics.

By implementing and analyzing the KNN algorithm for heart disease prediction, this study
contributes to the ongoing efforts to integrate AI and machine learning into the medical field,
ultimately leading to better patient outcomes.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

10
 4. METHODOLOGICAL DETAILS

4.1 SYSTEM ARCHITECTURE

Figure 4.1

Any project is basically divided into many groups for easy understanding and coding. This
paper consists of four which the application runs on. They are namely

• 4.1.1. User registration

• 4.1.2. Questionnaires
• 4.1.3. Checking the probability
• 4.1.4. Generating the report

4.1.1 USER REGISTRATION

Logging in, (or logging on or signing in or signing on), is the process by which an individual
gains access to a computer system by identifying and authenticating themselves. The user
credentials are typically some form of “username” and a matching “password”, and these
credentials themselves are sometimes referred to as a login.

4.1.2. QUESTIONNAIRES

Here the user feed the values in the application form, he/she fills up each and every details in
the form. All these details gets saved in the server and details and from that we can extract the
PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25
11
features of the disease. The entered details are matched with the datasets which are saved in
the database.

4.1.3. CHECKING THE PROBABILITY

After matching the details with the datasets it checks for the disease symptoms. One feature
may match with different disease. So, it’s necessary to check each and every matched details
in order to predict the correct disease.

4.1.4. GENERATING THE REPORT

That registered data is collected in a database by using machine learning techniques like data
collection techniques and when he went to check about his health condition the collected values
or data that has been stored in the database is been extracted by using some feature extraction
techniques. When data is extracted, it under goes certain processes and therefore finally a
disease is predicted and a report is generated. This is the overview of the heart disease
prediction system using machine learning techniques.

4.1.5 K-NEAREST NEIGHBORS (KNN)

K-Nearest Neighbors (KNN) is a supervised machine learning algorithm used for classification
and regression tasks. In heart disease prediction, KNN helps in classifying a patient as high-
risk or low-risk based on their symptoms, medical history, and lifestyle factors.

How KNN Works in Heart Disease Prediction

• Data Collection: Patient health records, including symptoms, blood pressure,

cholesterol levels, heart rate, and other risk factors, are collected in a dataset.

• Feature Selection & Normalization: The collected data is preprocessed, removing

missing values and scaling the features to ensure uniformity.

• Distance Calculation: When a new patient's data is entered, KNN calculates the distance
between the new patient and existing patient records using methods like:

1. Euclidean Distance

2. Manhattan Distance

• Finding Nearest Neighbors: Based on the calculated distance, the algorithm identifies
the K nearest neighbors from the dataset.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

12
 • Classification: The majority class (e.g., "Heart Disease" or "No Heart Disease") among
the nearest neighbors determines the prediction.

Why KNN for Heart Disease Prediction?

• Simple & Easy to Implement: Requires minimal training, making it efficient.

• No Assumptions on Data Distribution: Works well even with non-linear data.

• Adaptable to Real-Time Data: Can continuously update predictions as new data is

added.

Challenges of KNN in Heart Disease Prediction

• Computationally Expensive: As the dataset grows, distance calculations become

slower.

• Sensitive to Irrelevant Features: Requires careful feature selection to improve accuracy.

• Imbalanced Data Issues: If one class is significantly larger, KNN may be biased towards
it.

4.3 TIMELINE CHART

The Timeline Chart represents the project's development schedule, showing key milestones
and deadlines. It ensures that the project stays on track by clearly defining task allocations and
completion times.

Figure 4.2

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

13
4.4 DFD LEVEL – 0
The DFD Level-0 provides a high-level overview, treating the Heart Disease Prediction System
as a single process, interacting with:
• User (Patient) – Inputs health data.
• Database – Stores and retrieves user records.

Figure 4.3

4.5 DFD LEVEL 1

The DFD Level-1 breaks the system into sub-processes:
• Collecting user health data.
• Processing it using the KNN algorithm.
• Storing results in the database.

Figure 4.4
PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25
14
4.6 DFD LEVEL 2
The DFD Level-2 further expands internal processes:
• Accepting user inputs (age, cholesterol, blood pressure, etc.).
• Running the KNN model to classify the risk level.
• Displaying the prediction result to the user.

Figure 4.5
4.7 USE CASE DIAGRAM
The Use Case Diagram represents the interaction between:
• User (Patient) – Enters medical data and receives prediction.
• System – Processes data and provides results.

Figure 4.6
PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25
15
4.8 SEQUENCE DIAGRAM
The Sequence Diagram illustrates the flow of system operations:
1. User enters medical parameters.
2. System validates input.
3. KNN algorithm processes data.
4. Result is displayed to the user.

Figure 4.7
4.9 ER DIAGRAM
The ER Diagram represents the database structure, showing key entities:
• User (stores patient details).
• Medical Data (stores input health parameters).
• Prediction Result (stores risk classification output).

Figure 4.8

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

16
4.10 CLASS DIAGRAM
The Class Diagram shows the object-oriented structure of the system:
• User Class (attributes: name, age, health data).
• KNN Model Class (methods for classification).
• Database Class (stores input and output records).

Figure 4.9
4.11 COMPONENT DIAGRAM
The Component Diagram visualizes how software modules interact:
• Frontend UI – User interface for input/output.
• Processing Module – Executes KNN algorithm.
• Database – Stores medical records and predictions.

Figure 4.10
PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25
17
 [Link]

Figure 5.1
This project is a Disease Prediction System named HealthGuard™, designed to predict the
likelihood of heart disease, diabetes, and hypertension based on user-provided health
parameters. The system takes input from the user in the form of the following data:

User Input Parameters:

1. Full Name – The patient's name for identification.

2. Age (years) – The age of the patient.

3. Sex (0=Female, 1=Male) – The gender of the patient.

4. Chest Pain Type (0-4) – Different types of chest pain experienced.

5. Resting BP (mmHg) – The patient’s resting blood pressure.

6. Cholesterol (mg/dL) – The cholesterol level in the blood.

7. Fasting Sugar (0=No, 1=Yes) – Indicates whether fasting blood sugar is high.

8. Resting ECG (0-2) – Results from an electrocardiogram test.

9. Max Heart Rate – The maximum heart rate achieved.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

18
 Figure 5.2

This is a comprehensive medical report generated for the user, providing an in-depth analysis
of their health status. It includes the patient’s demographic and clinical information, a summary
of test results, and detailed findings on detected diseases such as heart disease and
hypertension. Additionally, the report offers risk assessments, potential contributing factors,
and recommended courses of action, including lifestyle modifications and medical
interventions. A general health advisory section is also included, outlining preventive
measures, dietary recommendations, and the importance of regular medical check-ups to
manage and mitigate potential health risks.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

19
 [Link]

In this project, we implemented a heart disease prediction system using machine

learning techniques, specifically the K-Nearest Neighbors (KNN) algorithm. The study
aimed to enhance early detection, improve diagnostic accuracy, and provide a decision-
support system for healthcare professionals.
The system successfully analyzes patient data by comparing it with stored datasets and
predicting the likelihood of heart disease. Through preprocessing, feature extraction,
and classification, the model identifies patterns that assist in diagnosing patients at risk.
The use of KNN in this context proved effective, demonstrating its ability to handle
medical datasets efficiently while maintaining a balance between accuracy and
computational complexity.
The project highlights the potential of machine learning in healthcare, contributing to
automated diagnosis and decision-making. While the model performs well, further
improvements can be made by integrating additional machine learning techniques,
optimizing feature selection, and incorporating real-time health monitoring through IoT
devices.
Overall, this study reinforces the importance of AI-driven healthcare solutions, paving
the way for future advancements in predictive analytics for heart disease and other
medical conditions.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

20
 [Link]

1. Amin, S. U., Agarwal, K., & Beg, R. (2019). "Genetic Neural Network Based Data
Mining in Prediction of Heart Disease Using Risk Factors." Proceedings of the
International Conference on Computing and Communication Technologies, IEEE.
2. Dua, D., & Graff, C. (2019). UCI Machine Learning Repository - Heart Disease
Dataset. University of California, Irvine, School of Information and Computer
Sciences.
3. Kumar, R., Kumar, R., & Bharti, R. (2021). "Comparative Analysis of Machine
Learning Algorithms for Heart Disease Prediction." International Journal of Advanced
Computer Science and Applications (IJACSA), 12(5), 45-52.
4. Gupta, A., & Sharma, M. (2020). "Performance Evaluation of KNN and SVM in Heart
Disease Prediction." International Journal of Scientific & Technology Research, 9(2),
78-84.

PGMCOE, DEPARTMENT OF ARTIFICIAL INTELLIGENCE AND DATA SCIENCE 2024-25

21