INTRODUCTION

The Core topic is prediction using machine learning techniques.

Machine learning is extensively used nowadays in numerous business

applications like e-commerce and many more. Prediction is one of an

area where the machine learning used, our subject is about the

prediction of heart disease by processing patient’s dataset and a data

of patients to whom we need to predict the chance of occurrence of

heart disease. Python is the most usual programming language having

several libraries which are used in this project with a machine

learning model. Machine learning is a subset model of an artificial

intelligence network which uses complex algorithms and deep

learning neural networks. The human body is made up of numerous

organs all of which have their functions. The heart is one such organ

that pumps blood all over the body and if it does not do so, the human
body can have fatal conditions. One of the key reasons for mortality

today is having heart disease.

Therefore `it becomes necessary to ensure that our cardiovascular

system or any other system in the human body for that matter should

remain healthy. Unfortunately, people around the world are suffering

from cardiovascular diseases. Any technology that can help detect

these diseases before much damage is done will shows that it is

helpful in saving peoples money and further significantly their life.

Data mining techniques can be beneficial in predicting heart diseases.

Analytical models can be made by finding earlier

unknown patterns and trends in databases and using the obtained

information. Machine learning is a technology that can help to

accomplish a diagnosis of heart disease before much damage occurs

to a person.
 LITERATURE SURVEY
1. Senthil Kumar Mohan, (Title) Effective Heart Disease

Prediction ML techniques ,(Objective) Objective is to find

critical includes by applying ML Algorithms , bringing about

improving the exactness in the prediction.

2. Sonam Nikhar,(Title)Prediction of heart disease using ml,

(objective) to give a point by point portrayal pf naïve bayes and

decision tree classifier.

3. VV Ramalingam ,(Title) heart disease prediction using ml,

(objective) The paper presents a survey of various models based

on algorithm that found very popular among the researchers.

4. Vijeta Sharma ,(Title) Heart Disease prediction using Ml

techniques,(Objective) find the correlations between the

different attributes available in the dataset with the help of

standard ML.
5. Dr M Kavitha ,(Title) heart disease prediction using hybrid

machine learning model, (objective) in this the Cleveland heart

disease dataset and machine learning algorithm are used for

prediction.
 METHODOLOGY AND
ANALYSIS
Data Collection
The data has collected from an online platform “KAGGLE”.

It contains 14 features and the dataset is the information or tool that is

important for any type of study or project.

Data Processing

I. Separation of target data and feature data as training and testing

data.

II. Scrambling the values in the data to be valued between 0 and 1

and scale all the values before training the Machine Learning

models.

Applying Algorithm

I. Equating 4-machine learning algorithms such as

SVM ,Decision Tree ,Random Forest and Logistic

Regression to get the improved accurateness to which highest

parameter may cause disease.

II. For every algorithm we find accuracy and also compare it in

a graph .In python there is a simple way to create any kind of

algorithm in which simple and short code is easier to predict

accuracy.
 MACHINE LEARNING ALGORITHM
The Algorithm used in this project are highly helpful to predict the

correct result to detect heart disease in which factors that cause

disease can be detected. The Following algorithms have been

assembled in this project.

1.RANDOM FOREST CLASSIFIER- A random forest classifier is

a powerful tool in the machine learning library. With the classifier ,

we will able to increase accuracy, and training time should be a

smaller amount. Primarily, we have to build a model and splitting

variables into training and test set. After that train the dependent

variables and predict the answer. By using the random forest classifier

the accuracy is 80%.

2.DECISION TREE CLASSIFIER- In this algorithm pre processing

is initially created by dividing the data into training data and test data.

Feature scaling can be done by normalizing the values before the

forecast.
Import decision tree classification into training sets of determined and

independent variables using guinea index criteria to predict the

accuracy or response of a test set. The accuracy obtained with the

algorithm is 80%.

3.SUPPORT VECTOR MACHINE- SVM is also one of the

classification algorithms in machine learning in which improved

accuracy can be predicted . As compared to other algorithms. It is

much better for expectedly predicting accuracy.in our prediction the

predicted accuracy is 82%.

4.K-NEAREST NEIGHBOUR- KNN is a managed classifier that

carries out an observation from within a test set to predict

classification labels. KNN is one of the classification methods used

whenever there is a classification. It has several assumptions includes

the dataset has slight noise, is labeled and it should covers relevant

features. By applying KNN in huge datasets takes a long time to

process. The accuracy grown with this algorithm is

 PROPOSED SYSTEM

ML
ALGORITHM

DATA FEATURE
MACHINE
DATASET PREPREOCES EXTRACTI
LEARNING MODEL
SING ON

DATA CLASSIFIER
RESULT
CLASSIFICATION
After confirming that the data is balanced the connection between the
data is found out and is plotted as a heat map.
From the above heatmap, we can understand that Chest Pain and
Target have a positive correlation. It means that whose has a large risk
of chest pain results in a greater chance to have heart disease. In
addition to chest pain, thalach, slope and resting have a positive
correlation with the target.
Then exercise-induced angina(exang) and the target have a negative
correlation which means when we exercise, the heart requires more
blood, but narrowed arteries slow down the blood flow. In addition to
ca, old peak also have a negative correlation with the target.
We have used 4 different algorithms to get the best accuracy .Decision
Tree Algorithm achieved the best accuracy of 91.9%.
 CONCLUSION

Applying all 4 algorithms to get best result in heart disease prediction.

Logistic Regression, Decision Tree, Random Forest Classifier and
Support Vector Machine all shows different accuracy. But out of all
Decision Tree gives the highest accuracy
of 92%.After applying various algorithms , it can be said that machine
learning is proving to be extremely valuable in predicting
heart disease which is one of the most prominent problems of society
in today’s world. As more and more work is being done in the field of
machine learning, new methods may soon come to make machine
learning more helpful in the field of healthcare. The
algorithms used in this experiment have performed well using the
available attributes. The conclusion can be finally drawn that machine
learning can reduce the damage done to a person physically and
mentally by predicting heart disease.
 REFERENCES

 Senthil Kumar Mohan (2019).Effective Heart Disease

Prediction Using Hybrid Machine Learning Techniques.

 Sonam Nikhar(2016).Prediction of Heart Disease using Machine

Learning Algorithms. International Journal of Advanced
Engineering, Management and Science(IJAEMS).

 V.V Ramalingam (2018).Heart Disease prediction using

machine learning techniques. International Journal of
Engineering & Technology.

 Vijeta Sharma(2020).Heart Disease Prediction using Machine

Learning Techniques.2nd International Conference on Advances
Computing ,Communication Control and
Networking(ICACCCN).

 Dr . M .Kavitha(2021).Heart Disease Prediction using Hybrid

machine learning Model. Proceedings of the Sixth International
Conference on Incentive Computation Technologies.