International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072

Heart Disease Prediction Using Machine Learning Techniques

Galla Siva Sai Bindhika1, Munaga Meghana2, Manchuri Sathvika Reddy3, Rajalakshmi4
1,2,3Student, Department of Computer Science, R.M.D. Engineering College
4 Assistant professor of Computer Science, R.M.D. Engineering College
----------------------------------------------------------------------------------***----------------------------------------------------------------------------
Abstract – impacting the society towards health. The main
concept is to identify the age group and heart
rate using the Random forest algorithm. Our
Heart disease is one of the most significant project tells how the heart rate and condition is
problem that is arising in the world today. estimated based on the inputs such as blood
Cardiovascular disease prediction is a critical pressure and many more being provided by the
challenge in the area of clinical data analysis. user to a system. This is being much better way
Hybrid Machine learning (ML) has been showing when it comes with others algorithms the
an effective assistance in making decisions and implementation of RFA gives the better
predictions from the large quantity of data experience and provide accurate result. This
produced by the healthcare industries and helps in early prediction of the disease and is
hospitals. We have also seen ML techniques used in many ways, where as it is being provided
being used in recent developments in different with the input, in order to find the heart rate
areas of the Internet of Things (IoT). Various based on the health condition.
studies give only a glimpse in predicting heart
disease with ML techniques. In this paper, we 2.EXISTING SYSTEM
propose a narrative method that aims at finding
significant features by applying machine learning In this system, the input details are obtained
techniques that results in improving the from the patient. Then from the user inputs,
accuracy in the prediction of cardiovascular
using ML techniques heart disease is analyzed.
disease. The prediction model is proposed with
combinations of different features and several Now, the obtained results are compared with the
classification techniques. We produce an results of existing models within the same
enhanced performance level with an accuracy domain and found to be improved. The data of
level of 92% through the prediction model for heart disease patients collected from the UCI
heart disease with the hybrid random forest with laboratory is used to discover patterns with NN,
a linear model. DT, Support Vector machines SVM, and Naive
Bayes. The results are compared for
Key words – performance and accuracy with these
algorithms. The proposed hybrid method returns
Cardiovascular Disease Prediction, Machine results of 87% for F-measure, competing with
Learning Techniques, Random forest linear the other existing methods.
model.
2.1 DISADVANTAGES
1. INTRODUCTION
1. Prediction of cardiovascular disease
Now a days, heart disease prediction has results is not accurate.
been a major concept in recent world that is

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5272
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072

2. Data mining techniques does not help to

provide effective decision making.
3. Cannot handle enormous datasets for
patient records.

3. PROPOSED SYSTEM

After evaluating the results from the existing

methodologies, we have used python and pandas
operations to perform heart disease
classification for the data obtained from the UCI 4.2 Feature Selection and Reduction
repository. It provides an easy-to-use visual
representation of the dataset, working Among the 13 attributes of the data set,
environment and building the predictive two attributes pertaining to age and sex are used
analytics. ML process starts from a pre- to identify the personal information of the
processing data phase followed by feature patient. The remaining attributes are considered
selection based on data cleaning, classification of important as they contain vital clinical records.
modelling performance evaluation. Random Clinical records are vital to diagnosis and
forest technique is used to improve the accuracy learning the severity of heart disease.
of the result.

3.1 ADVANTAGES

1. Increased accuracy for effective heart

disease diagnosis.
2. Handles roughest(enormous) amount of
data using random forest algorithm and
feature selection.
3. Reduce the time complexity of doctors.
4. Cost effective for patients.

4. APPROACH
4.3 Classification Modelling
4.1 Data Pre-Processing
The clustering of datasets is done on the basis of
Heart disease data is pre-processed by using
the variables and criteria of Decision Tree (DT)
various collection of records. The dataset
features. Then, the classifiers are applied to each
contains a total of 303 patient records, where 6
clustered dataset in order to estimate its
records are with some missing values. Those 6
performance. The best performing models are
records have been removed from the dataset and
identified from the above results based on their
the remaining 297 patient records are used in
low rate of error.
pre-processing.

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5273
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072

4.3.1 DECISION TREES

For training samples of data D, the trees are

constructed based on entropy inputs. These
trees are simply constructed in a top down
recursive divide and conquer (DAC) approach.
Tree pruning is performed to remove the
irrelevant samples on D.

Entropy = −Xm j=1 pij log2 pij

Algorithm for Decision Tree-Based Partition

Require:
4.3.3 RANDOM FOREST
Input: D dataset – features with a target class
for ∀features do This ensemble classifier builds several
for Each sample decision trees and incorporates them to get the
do Execute the Decision Tree algorithm best result.
end for Identify the feature space f1, f2, . . . , fx of
dataset UCI. For tree learning, it mainly applies bootstrap
end for Obtain the total number of leaf nodes aggregating or bagging.
l1,l2,l3,. . . ,
ln with its constraints Split the dataset D into For a given data, X = {x1, x2, x3, ..., xn} with
d1, d2, d3,. . . , dn responses Y = {x1, x2, x3, ..., xn} which repeats
based on the leaf nodes constraints. the bagging from b = 1 to B
Output: Partition datasets d1, d2, d3,.
Generating the input using python and
random forest classification

4.3.2 LANGUAGE MODEL

For given input features (xi, yi) with input

vector xi of data D the linear form of solution
f (x) = mx+b equation is solved by
subsequent parameters:
4.3.4 SUPPORT VECTOR MACHINE

Let the training samples having dataset Data

m=P
= {yi, xi}; i = 1, 2, . . . , n where xi ∈ R n represent
the i th vector and yi ∈ R n represent the target

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5274
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072

item. The linear SVM finds the optimal 5. Performance Measures

hyperplane of
Several standard performance metrics such as
accuracy. precision and error in classification
have been considered for the computation of
performance efficiency of this model.

the form f (x) = w T x + b where w is a

dimensional coefficient vector and b is a offset.
This is done by solving the subsequent
optimization problem:

Minw,b, ξi 1 2 w 2 + C Xn i=1 ξi s.t. yi , w T

xi + b ≥ 1 − ξi, ξi ≥ 0, ∀i ∈ {1, 2, . . . ,}

6. Conclusion
In this paper, we proposed a method for heart
disease prediction using machine learning
techniques, these results showed a great
accuracy standard for producing a better
estimation result. By introducing new proposed
Random forest classification, we find the
problem of prediction rate without equipment
Comparing the obtained input results using and propose an approach to estimate the heart
Support Vector Machine and Language model rate and condition. Sample results of heartrate
classifier: are to be taken at different stages of the same
subjects, we find the information from the above
input via ML Techniques. Firstly, we introduced
a support vector classifier based on datasets.

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5275
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 07 Issue: 04 | Apr 2020 www.irjet.net p-ISSN: 2395-0072

7. REFERENCES [8] H. A. Esfahani and M. Ghazanfari,

‘‘Cardiovascular disease detection using a new
[1]A. S. Abdullah and R. R. Rajalaxmi, ‘‘A data ensemble classifier,’’ in Proc. IEEE 4th Int. Conf.
mining model for predicting the coronary heart Knowl.Based Eng. Innov. (KBEI), Dec. 2017, pp.
disease using random forest classifier,’’ in Proc. 1011–1014.
Int. Conf. Recent Trends Comput. Methods,
Commun. Controls, Apr. 2012, pp. 22–25. [9]F. Dammak, L. Baccour, and A. M. Alimi, ‘‘The
impact ofcriterion weights techniques in TOPSIS
[2]A. H. Alkeshuosh, M. Z. Moghadam, I. Al method of multi-criteria decision making in crisp
Mansoori, and M. Abdar, ‘‘Using PSO algorithm and intuitionistic fuzzy domains,’’ in Proc. IEEE
for producing best rules in diagnosis of heart Int. Conf. Fuzzy Syst. (FUZZ-IEEE), vol. 9, Aug.
disease,’’ in Proc. Int. Conf. Comput. Appl. (ICCA), 2015, pp. 1–8.
Sep. 2017, pp. 306– 311.
R. Das, I. Turkoglu, and A. Sengur, ‘‘Effective
[10]

[3]N. Al-milli, ‘‘Backpropogation neural network diagnosis of heart disease through neural
for prediction of heart disease,’’ J. Theor. networks ensembles,’’ Expert Syst. Appl., vol. 36,
Appl.Inf. Technol., vol. 56, no. 1, pp. 131–135, no. 4, pp. 7675–7680, May 2009. doi:
2013. 10.1016/j.eswa.2008.09.013.

[4] C.
A. Devi, S. P. Rajamhoana, K. Umamaheswari,
R. Kiruba, K. Karunya, and R. Deepika, ‘‘Analysis
of neural networks based heart disease
prediction system,’’ in Proc. 11th Int. Conf. Hum.
Syst. Interact. (HSI), Gdansk, Poland, Jul. 2018,
pp. 233–239.

P. K. Anooj, ‘‘Clinical decision support system:

[5]

Risk level prediction of heart disease using

weighted fuzzy rules,’’ J. King Saud Univ.-
Comput. Inf. Sci.,vol.24,no.1,pp.27–
40,Jan.2012.doi:10.1016/j.jksuci.2011.09.002.

[6] L.
Baccour, ‘‘Amended fused TOPSIS-VIKOR for
classification (ATOVIC) applied to some UCI data
sets,’’ Expert Syst. Appl., vol. 99, pp. 115–125,
Jun. 2018. doi: 10.1016/j.eswa.2018.01.025.

[7]C.-A.Cheng and H.-W. Chiu, ‘‘An artificial

neural network model for the evaluation of
carotid artery stenting prognosis using a
national-wide database,’’ in Proc. 39th Annu. Int.
Conf. IEEE Eng. Med. Biol. Soc. (EMBC), Jul. 2017,
pp. 2566–2569.

© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 5276