GKM College of Engineering and Technology

HEART DISEASE PREDICTION

USING K NN Algorithm

Team Members: Guide:

P.Prasanth (410819104020) Mrs. Dayana
Assistant Professor, CSE
 Objective

Preventing heart disease is important. Good data-driven systems for predicting

heart disease can improve the entire research and prevention process, making sure
that more people can live healthy lives.

 Early prediction of heart disease is essential to reduce the mortality rate. Data
mining provides a user-oriented approach to extract novel and uncovered patterns
in the data set. Data mining is to extract useful knowledge within medical data for
medical diagnosis
 Abstract
Heart disease commonly occurring disease and is the major cause of sudden death
nowadays. This disease attacks the persons instantly. Most of the people do not aware of the
symptoms of heart disease. Timely attention and proper diagnosis of heart disease will reduce the
mortality rate. Medical data mining is to explore hidden pattern from the data sets. Supervised
algorithms are used for the early prediction of heart disease. Nearest Neighbor (KNN) is the widely
used lazy classification algorithm. KNN is the most popular, effective and efficient algorithm used for
pattern recognition. Medical data sets contain a large number of features. The Performance of the
classifier will be reduced if the data sets contain noisy features. Feature subset selection is proposed
to solve this problem. Particle Swarm Optimization (PSO) is an Evolutionary Computation (EC)
technique used for feature selection. PSO are computationally inexpensive and converges quickly. In
this we apply KNN and PSO for prediction of heart disease. Experimental results show that the
algorithm performs very well with 100% accuracy with PSO as feature selection.
 Literature survey
SI.NO Author Name Title of Paper Remarks

1 Vijeta Sharma , Heart Disease Prediction using  Use of benchmark dataset which
Shrinkhala Yadav, Machine Learning Techniques consists of only 14 parameters.
Manjari Gupta [May 2020]  Using random forest ,which
reduces the accuracy of prediction
for complex problems

2 Archana Singh, Heart Disease Prediction using  SVM(Support Vector Machine)

Rakesh Kumar Machine Learning  Navie Bayes and Decision tree
Algorithms[June 2020] used as development model
SI.NO Author Name Title of Paper Remarks
3 M. Snehith Raja, Machine Learning Based Heart Disease  Prediction is mostly based on old
M. Anurag, Prediction System[April 2021] patient record.
Ch. Prachetan  Data set is feed in format of CSV
Reddy, file which will reduce the accuracy.
Nageswara Rao
Sirisala

4 Reldean Williams, Heart Disease Prediction using  Use of Decision tree- That increases
Thokozani Machine Learning Techniques [Dec the processing time
Shongwe, 2021]  Use of logistic regression – which
Ali N. Hasan, fails to predict a continuous
Vikash Rameshar outcome.
SI.NO Author Name Title Remarks
Sanchayita Dhar,Pritha A Hybrid Machine Learning Uses Random Forest ,decision tree and
5 Datta , Ankur Biswas, Approach for Prediction of Naïve Bayes that integrates a prediction
Tanusree Dey, Krishna Heart Diseases [2018] system that is capable to envisage heart
Roy diseases based on measurements extracted
from ERIC laboratory.
 Problem Identification

A report by Indian Council of Medical Research(ICMR) said that cardiovascular

ailments account for nearly 17 percent of all deaths in South Asian country. On average,
around 10 million people die in India every year.
At present , use of machine learning for prediction, identification and recognition of
various real time problems which increases the accuracy rate and also helps to handle
large amount of data.
In order to make the diagnosis process in a much efficient and advanced manner , in that
way for heart disease prediction the proposed system acts as one of the efficient
machine learning methodology.
 Introduction

The heart disease a major concern to be dealt with. But it is difficult to identify
heart disease because of several contributory risk factors such as diabetes, high
blood pressure, high cholesterol, abnormal pulse rate, and many other factors. Due
to such constraints, scientists have turned towards modern approaches like Data
Mining and Machine Learning for predicting the disease.
Machine learning (ML) proves to be effective in assisting in making decisions and
predictions from the large quantity of data produced by the healthcare industry.
 System Architecture
KNN Algorithm ,Random
Forest algorithm

Apply pre-processing Preprocessed Train Set Apply computational

technique Dataset intelligent technology

Test Set
Clinical Model Evaluation
Dataset

Accuracy , sensitivity,
Specification , Precision ,
MCC
 Modules

1. Data Preprocessing

2. Feature Selection

3. K-NN model development

4. Model evaluation

5. Deployment
 1.Data Preprocessing
The first module involves preparing the dataset for the K-NN algorithm by :

 Cleaning the data

 Transforming the data

 Normalizing the data.

The steps involved in data preprocessing are:

 Handling missing values

 Dealing with outliers

 Scaling the data

 2.Feature Selection

In this module, we identify the most relevant features or variables that
can influence the prediction of heart diseases.
Techniques involved:

 Correlation analysis

 Feature importance ranking

 3.K-NN Model Development

Here, we develop the K-NN algorithm by selecting the appropriate

value of k.
Testing the model on a training dataset.

Module also includes hyperparameter tuning to optimize the

performance of the model.
 4.Model Evaluation
In this module we evaluate the performance of the K-NN model
using various metrics such as accuracy, precision, recall and F1
score.

The module compares the performance of the K-NN model with

other machine learning algorithm.
 5.Deployment

This final module involves deploying the K-NN model in a

production environment, which will be an web application.

This module also includes a user-friendly interface for users to input

their health data and receive heart disease predictions from the K-NN
model.
 Methodology
KNN Algorithm:
 K-Nearest Neighbour is a supervised machine learning algorithm
that uses labelled input data set to predict the output of the data points.
 KNN algorithm is used for classification and regression. It is a
versatile algorithm also used for imputing missing values and
resampling datasets.
 As the name suggests it considers K Nearest Neighbors (Data
points) to predict the class or continuous value for the new data.
 In this we have proposed a Heart disease prediction which makes
use of KNN instead of random forest which increases the accuracy rate
from 85% to around approx. of 96%.
 Result with Discussion

In this project , we developed a model using K-NN algorithm for the
prediction of cardiovascular disease.
The model was trained on a dataset of 70,000 patient records, which
were collected from various hospitals and clinics.
The dataset contained a total of 16 features, including age, gender,
blood pressure, cholesterol level and smoking status.
 Publication

• Abstract Submission in the “ Third international conference on

Recent Trends in Science and Technology” organised by GKM
College of Engineering and Technology on 28.3.2020.

• Paper Abstract submission date : 13.3.2020

• Conference website : www.icrtst.com

11/3/2019 Project title 18

 Reference
 S.Pouriyeh, S.Vahid, G.Sannino, G.De Pietro, H.Arabnia and J.Gutierrez, “A
comprehensive investigation and comparison of Machine Learning Techniques in
domain of heart disease”,2017 IEEE Symposium on Computers and Communications.

S.Dhar, K.Roy, T. Dey, P. Datta and A.Biswas, “A Hybrid Machine Learning

Approach for Prediction of Heart Disease”,2018 4th ICCCA.

C.Raju, E.Philipsy, S.Chacko, L. Padma Suresh and S.Deepa Rajan, “A survey on

Predicting Heart Disease using Data Mining Techniques” 2018 Conference on
Emerging Devices and Smart Systems.