International Journal of Advances in Applied Sciences (IJAAS)

Vol. 11, No. 2, June 2022, pp. 158~167

ISSN: 2252-8814, DOI: 10.11591/ijaas.v11.i2.pp158-167  158

Heart disease classification using various heuristic algorithms

Arif Ullah1, Shakeel Ahmad Khan2, Tanweer Alam3, Shkurte Luma-Osmani4, Mahanz Sadie5
1
Department of Computing, Riphah International University, Faisalabad, Pakistan
2
Directorate of Higher Education, Government of Khyber Pakhtunkhwa Peshawar, Khyber Pakhtunkhwa, Pakistan
3
Department of Computer Science, Islamic University of Medina, Medina, Saudi Arabia
4
Faculty of Natural Sciences and Mathematics, University of Tetovo, Tetovo, North Macedonia
5
Department of Computer Engineering, Ramhormoz Branch, Islamic Azad University, Ramhormoz, Iran

Article Info ABSTRACT

Article history: In the health sector, the computer-aided diagnosis (CAD) system is a rapidly
growing technology because medical diagnostic systems make a huge
Received Apr 21, 2021 change as compared to the traditional system. Now a day huge availability of
Revised Dec 14, 2021 medical data and it needs a proper system to extract them into useful
Accepted Apr 20, 2022 knowledge. Heart disease accounts to be the leading cause of death
worldwide. Heuristic algorithms have been exposed to be operative in
supporting making decisions and classification from the large quantity of
Keywords: data produced by the healthcare sector. Classification is a prevailing
heuristic approach which is commonly used for classification purpose some
Algorithm heuristic algorithm predicts accurate result according to the marks whereas
Classification some others exhibit limited accuracy. This paper is used to categorize the
Diseases attendance of heart disease with a compact number of aspects. Original, 13
Healthcare attributes are involved in classifying heart disease. A reasonable analysis of
Prediction these techniques was done to conclude how the cooperative techniques can
be applied for improving prediction accuracy in heart disease. Four main
classifiers used to construct heart disease prediction based on the
experimental results demonstrate that support vector machine, artificial bee
colony (ABC), bat algorithm, and memory-based learner (MBL) provide
efficient results. The accuracy differs between 13 features and 8 features in
the training dataset is 1.9% and in the validation, dataset is 0.92% of vector
machine which is the most accurate heuristic algorithm.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Arif Ullah
Department of Computing, Riphah International University
Faisalabad, Punjab, 44000, Faisalabad, Pakistan
Email, [email protected]

1. INTRODUCTION
Medical data design in healthcare is one of the important and complicated tasks because as a result,
it must show accuracy and efficiency in its performance. Computer-aided diagnosis (CAD) becomes a very
useful tool that attempts to solve real-world health problems in the diagnosis and treatment of disease [1]. In
present years, heuristic systems have found their significant hold in every field comprising health care.
Medical history data includes a number of tests prescribed to diagnose a specific disease [2]. Clinical
databases are elements of the domain where the route of heuristic technique has advanced into an inevitable
aspect due to the measured incline of medical and clinical research data. It is possible for the healthcare
industries to gain an advantage of the heuristic technique by employing the same as an intelligent
investigative tool. The report of heuristic technique is high accuracy and absorbing rate in this field.
Normally, many health care organizations are encrustation a major defy to offer high-quality provisions, like

Journal homepage: http://ijaas.iaescore.com

Int J Adv Appl Sci ISSN: 2252-8814  159

analyzing patients appropriately and organizing treatment at sensible costs [3], [4]. The heuristic technique
has been widely used to mine information from medical databases. In the heuristic technique, a classification
is a form of learning that can be used to design models describing important data classes. Using those
heuristic techniques can provision researchers or physicians in making medical verdicts and they can answer
significant and connected questions regarding health care. It is problematic to recognize heart disease
because of several contributory risk issues such as diabetes, high blood pressure, high cholesterol, abnormal
pulse rate, and many other factors [5], [6]. The rest of this paper is organized as: The necessary background
for the heart disease is discussed in section 2. Section 3, presents the related work in this section. In section 4,
the statistical analysis and limitations are presented. In Section 5, we present the outstanding issues of
heuristic algorithms for heart disease prediction and their result. Section 6, presents the conclusion. For the
comfort of readers, we provided a list of the most frequently used acronyms in the paper are mentioned in
Table 1.

Table 1. List abbreviations

Abbreviations Meaning Abbreviations Meaning
HD Heart diseases Bat Bat algorithm
PH Predication approach ABC Artificial bee colony
MBL Memory-Based Learner DP Deep learning
SVM Support vector machine M.SVM Modified support vector machine
WHO World Health Organization UCI University of California Irvine
CAD Computer-aided diagnosis MBL Memory-based learner

2. DISEASES
The disease is a particular term that is used for the abnormal condition, or this term is used in
healthcare conditions that part of an organism. Various diseases are known by their symptoms, signs, and
knowledge and therefore it is necessary to optimize disease forecaster. According to a survey by World
Health Organization (WHO), heart disease has been the main cause of death in the world. Heart attack in
different countries due to exertion, work overload, mental stress, food, and so on the treatment and diagnosis
is complicated therefore it is an important step to design a systematic system for diagnoses [7], [8]. The main
prevention of this disease is to make an early prediction. Heart disease risk factors the main elements which
make a personal risk factor for this disease are. Age factors: Normally men face heart attack at age of 65 and
women face this kind of attack after. Climacteric but as compared to women, men face more heart attacks.
Family history; normally it is seen that the family faces this kind of problem and it inherits from family to
family heart disease cause. Smoking factor a chemical in tobacco smoke that developed blood clots and also
increase the cause of heart attacks now a day it is more in the young generation due to the use of smoke
weight; if the body increases then it makes the chance of heart disease. These people who carry extra body
facts give chance to these diseases. To reduce this kind of numerous problems any kind of factor can be used
[9], [10].

3. RELATED WORK
Different researchers have proposed several heuristic algorithms and techniques to predict heart
disease accurately some of them are. Agostino et al. [11], designed a neural network algorithm known as
self-organizing piecewise aggregate approximate (SOPAA). This algorithm is used for electrocardiogram
(ECG) signal classification and performed to diagnose heart disease. According to their result, 97% better
than other techniques used. Researchers implement a genetic algorithm in a neural network for heart disease
prediction. They use 12 parameters such as sex, age, blood, and cholesterol. The algorithms calculate the
number of hidden layer nodes and architecture to calculate the parameter result. According to their result, it
gives up to 98% accuracy16. According to [12], design a scheme that is known as a prototype intelligent
heart disease prediction system (IHDPS) using deep learning technique, namely decision trees. ID3. By using
medical profiles of patients such as age, gender, blood pressure and blood sugar, chest pain, and an
electrocardiogram (ECG) graph, it can predict the likelihood of patients getting a heart disease or not and
make 82% effectually working in necessity. Concurring to [13], design a multilayer perceptron model for
analytical detection of heart disease severity based on various factors. The author also organizes a novel
principle attribute analysis to understand the direction of the attributes emotional the results. The final effect
of this work is to explore the heart disease severity based on the planned multilayer perceptron model
conferring the result it 97% accurate.
Ullah et al. [14] designed a scheme known as a decision tree and grabbing algorithm used for the
reduction of blood and oxygen supply to the heart leads to heart disease. Used disparate attributes to associate

Heart disease classification using various heuristic algorithms (Arif Ullah)

160  ISSN: 2252-8814

with other algorithms and the result shows a 96% increment in this scheme [15], proposed for the diagnosis
of coronary artery disease the decision tree is used to select the most significant attributes, and the output is
converted into crisp if-then rules. The crisp sets of rules are transformed into the fuzzy rules and these rules
constitute the fuzzy rule base. The performance of the proposed system is analyzed using various parameters
like classification accuracy, sensitivity, and specificity and it is observed that that this system achieves better
accuracy about 98% than the existing systems [16]. Author applied and compared several data mining
techniques to predict heart diseases. They used models based upon five algorithms including C5.0, neural
network, support vector machine (SVM), k-nearest neighborhood (KNN), and logistic regression. C5.0
decision tree was able to build a model with greatest accuracy of 93.02% whereas KNN, SVM, neural
network had the accuracy of 88.37%, 86.05%, and 80.23% respectively [17]. Summary of the related work
are mention in Table 2.

Table 2. Related work summary

Heuristic Algorithms Year Disease Dataset Simulation Accuracy rate
Bayes Net/ FT 84.5% 2015 Coronary artery disease UCI WEKA 84.5%
Naive Bayes/SVM
Naive Bayes/ J48 2015 Heart Disease UCI WEKA 85.1
Bagging/ SVM 2018 Heart Disease UCI WEKA 84.2
UCI 2019 Heart disease UCI WEKA 84.95
ANN 2019 Dengue Disease UCI WEKA 85.56
MFNN 2019 Dengue Disease UCI WEKA 84.56
C4.5 2019 Dengue Disease UCI WEKA 84.56
Feed forward NN with Back propagation 2019 Dengue Disease UCI WEKA 84.56
CART 2019 Dengue Disease UCI WEKA 82.56
K Star 2019 Dengue Disease UCI WEKA 83.56

4. PROPOSED APPROACHES
In this section, we present all those classifiers of heuristic algorithms which are used for heart
disease prediction in this paper. Clevel and University of California Irvine (UCI) repository dataset for heart
disease classification process offers an easy-to-use visual representation of the dataset, working situation, and
building the analytical analytics. Applying the machine learning technique start from pre-processing phase
followed by post-processing steps.

5. CLASSIFIER EVALUATION MEASURES

In this research, the context selection technique is determined by using the weight of the lower-level
context technique. Rather than using the learning weights through heuristic algorithms. In this section, we
collect the information from each gain of the attribute as its weight. The basic concept of this study comes
from the SVM classification technique and it is used the concept of information gain as the criteria for
selecting an attribute process [18], [19]. SVM is justly used for story set selection so the technique applied
consists of computing the SVM for each field. SVM gives attribute X with respect to the class attribute Y is
the reduction in uncertainty about the value of Y when we know the value of X, I (Y; X). The uncertainty
about the value of Y when we know the value of X is given by the conditional entropy of Y given X, H (Y,
X) [20]. Figure 1 presents the proposed approaches.

Figure 1. Proposed approaches structure

Int J Adv Appl Sci, Vol. 11, No. 2, June 2022: 158-167
Int J Adv Appl Sci ISSN: 2252-8814  161

5.1. Support vector machine (SVM)

SVM has proven to be a supervised machine learning method, commonly used to provide data
analysis for regression and classification problems. Generally, this method is used as a classifier, based on
the principle of making a plot of n dimensions in space, where the value of each entity also presents the value
of the specified coordinate. Subsequently, find the ideal hyperplane that made it possible to differentiate the
two created classes [21], [22]. Like other machines, the SVM takes input, performs input manipulation, and
then provides an output, which can later categorize new examples. SVM classifier used for binary
classification of feature space by hyperplane-based linear separation of data. Normally used for nonlinear
regression and pattern classification purposes, SVM work on the following base rule [23], [24].

𝑆𝑔𝑛(𝑓(𝑥, 𝑤𝑏)
(1)
𝑓(𝑥, 𝑤, 𝑏) = (𝑤. 𝑥) + 𝑏

Where x is the example to be classified and the maximum margin with a hyper plane (𝑤, 𝑏) represent the
complex problem w within the specified constraints [25], [26].

yi(w.xi) + b)  1 (2)

( w.x) + b) = 0   R n , b R (3)

5.2. Memory-based learner (MBL)

The memory-based learner also known as instance-based this approach works on the nearest
neighbor that classifies the attributes based on training labels (K) as the neighbor in the feature space. A
distance function is used to identify the nearest neighbors. The type of distance function is contingent on the
data type of nominated attributes [27], [28].

𝑑(𝑥𝑖, 𝑥𝑗) = ∑𝑞𝛼𝑄(𝑋𝑖𝑞 − 𝑋𝑖𝑞 )2 + ∑𝐶𝜀𝑐 𝐿𝑐 (𝑋𝑖𝑐 − 𝑋𝑗𝑐 ) (4)

Where Lc represents the M × M matrix used to describe the distance between two categorical attributes, Let
Ni represents the set of k nearest neighbors for an instance xi having the distance of d, T
[29]–[33].

𝑉𝑖 (𝑡) = ∑𝐾𝜀𝑁𝑖 (𝑡, 𝑦𝑘)) (5)

5.3. Bat algorithm

Bat algorithm generates the new solution now it time to speed at which bat searches for prey in the
dimension present as dimension. These equations are used for the classification process.

(𝑡−1) (𝑡−1)
𝑉𝑖𝑑𝑡 = 𝑣𝑖𝑑 +  𝑓𝑖 (𝑥𝑖𝑑 − 𝑥𝑑∗ ) (6)

The given equation d presents the dimension of the search space 𝑉𝑖𝑑𝑡 and 𝑣𝑖𝑑
(𝑡−1)
denote the flight speed of the
(𝑡−1)
bat 𝑖 and 𝑡 − 1 respectively 𝑥𝑖𝑑 denote the position of bat 𝑖 at time 𝑡 − 1 and 𝑥𝑑∗ represent the best position
in the population up to the 𝑡 iteration so far [34], [35].

5.4. ABC algorithm

All the directions of the inhabitants of food causes, xm→’s, are initialized (m = 1...SN, SN:
population size) by scout bees and regulator parameters are set. Since each food source, xm→, is a solution
vector to the optimization problem, each xm→ vector holds n variables, (xmi, i = 1...n), which are to be
enhanced so as to minimize the classification purpose.

6. DATA PRE-PROCESSING
Dataset for this study Cleveland dataset (UCI, 2016) used which are under the control of UCI
machine learning repository used for heart disease dataset. The dataset consists of 303 instances of the patient
but 6 of them are conation missing values and Table 3 shows the dataset attributes with their definition [36],
[37].

Heart disease classification using various heuristic algorithms (Arif Ullah)

162  ISSN: 2252-8814

Table 3. Dataset features information and description

The domain of
Feature name Feature code Description
values (minimax)
Age AGE Age in yrea 20<age>80
Sex SEX Female =1, Male 0, 0,1
Resting blood pressure RBP mm Hg admitted at the hospital 94–200
Type of chest pain CPT 1. atypical angina 1
2. typical angina 2
3. asymptomatic 3
4. nonagonal pain 4
Fasting blood sugar >120 mg/dl FBS Fasting blood sugar >120 mg/dl (1 _ true; 0 _ false) 10
00
12
Serum cholesterol SCH In mg/dl 120–564
Resting electrocardiographic results RES having ST-T 1 1
hypertrophy 2 2
Maximum heart rate achieved MHR - 71–202
Exercise-induced angina EIA 1 _ yes 1
0 _ no 0
Old peak _ ST depression induced by OPK - 0–6.2
exercise relative to rest
Slope of the peak exercise ST PES 1. up sloping 1 1
segment 2. .at 2 2
3. down sloping 3
Number of major vessels (0–3) VCA - 12
colored by horoscope 3
Thallium scan THA 3.normal 3
6.£xed defect 6
7. reversible defect 7

7. RESULTS AND DISCUSSION

In this section, we present the result of different proposed classifiers in heuristic algorithms taking
UCI dataset. Before discussing the result, these results are evaluated by using evaluation measures which
areas:
In this study we evaluated the accuracy rate of prediction of proposed different models on heart
disease datasets are mentioned: i) Classification error: it is used for the measurement of incorrect
classification in the classification model which is measured as; ii) Classification accuracy: used to check the
overall result of the performance of different classification and it will be measured as; iii) Accuracy =
(TP + TN)
∗ 100 [38], [39]; iv) Sensitivity is the process which shows the positive fraction or confirms
(TP + FP + TN + FN)
that diagnostic test is positive and the test result for which process has find and it can be written as; v)
TP
Sensitivity/rec call /true positive rate =Sensitivity = * 100; vi) Specificity is diagnostic test is
(TP + FN)
TN TP
negative and person is healthy and can be present as: * 100; vii) Precision can be: * 100;
(TN + FP) (TP + FN)
viii) True Positive rate= TP / (TP + FN); ix) False Positive rate = FP / (FP + TN) [40]–[42].

8. EXPERIMENTS
In this study, Jupiter notebook software was used for heart disease prediction taking, cleverly and
heart disease dataset. Here, the diagnostic performance is evaluated in terms of accuracy, precision,
sensitivity, and specificity, taking different heuristic algorithms. The factors contributing to these are
discussed, see Table 4 and Table 5.
Figure 2 shows the result of different machine learning approaches in terms of accuracy
performance for heart disease prediction. Based on the result SVM classifier provide the best result and its
accuracy rate is 98.91% best case and worst case. Memory-based learner and its result was 93.62%. Table 6
presents the result of the proposed technique (Classification) results in terms of instances.

Table 4. Assessment of heuristic approach for heart disease prediction

Parameter Accuracy Sensitivity Specificity TP Rate FP Rate
Memory-Based Learner 81.08% 86.25% 75.82% 0.8625 0.2410
Bat 84.08% 86.25% 75.82% 0.8625 0.2410
ABC 79.05% 83.12% 74.26% 0.8312 0.2573
SVM 98.90% 97.45% 90.20% 0.92897 0.2175

Int J Adv Appl Sci, Vol. 11, No. 2, June 2022: 158-167
Int J Adv Appl Sci ISSN: 2252-8814  163

Table 5. Result of different classifier

Classifiers Performance
Memory-Based Learner 90.74%
Bat 93.62%
ABC 98.089%
SVM 98.91%

Figure 2. Rust of different classifier

Table 6. Proposed classifier rest

TP Rate FP Rate Precision Recall F-Measure ROC Area
Incorrectly Classified Instances 0.312 0.092 0.571 0.312 0.404 0.695
Correctly Classified Instances 0.908 0.688 0.571 0.908 0.834 0.695
Weighted Average 0.74 0.52 0.714 0.74 0.712 0.695

Table 6 shows the training section of proposed classifier approaches where different instances are
checked based on missing or incorrect and correct forms. To measure these instances different element is
used to check which are recalled, and the receiver operating characteristic (ROC) area-based on all parameter
the value of the correctly classified instance is better and proved good result by the different classifier during
the simulation process. Figure 3 present the result of all heuristic classifier that is used in the training and
testing section of the study. Taking different elements three main sections of the dataset are checked which
are correct instances and correct, average instance based on the result the different proposed approach
provides the good result and most of the instance is correct. Table 7 show the result of the classifier in term
of time. Table 7 present the one of most important parameters of classification which is time consumption
based on the table SVM and radial basis function network is the best case in term of time and accuracy used.
Mean absolute error (MAE) section for time measurement in the processing system. This is
primarily because it takes only a few milliseconds to calculate the accuracy purpose. Figure 4 present the
time consumption of different machine learning classifier during the processing system and Table 8 present
the overall time of all classifier taking different data range.

Figure 3. The result of the classifier

Heart disease classification using various heuristic algorithms (Arif Ullah)
164  ISSN: 2252-8814

Table 7. Time comparison of classifier

Classifiers Performance Time of construction MAE
Memory-Based Learner 90.5 0.02 0.044
Bat 91.2 0.08 0.00019
ABC 88.3 0.06 0.117
SVM 99.2 0.02 0.044

Figure 4. Time-based comparison of classifier

Figure 5 and Table 8 present the overall time consumption of the heuristic technique all the model is
good and improve the performance of every base classifier by providing accurate predictions of heart
diseases. The overall time is well based on the classifier result so are at the position we can say machine
learning approaches are good for a medical dataset.

Table 8. Overall time consumption

Classifiers Performance Time for construction
Memory-Based Learner 52.33% 609 ms
Bat 52% 719 ms
ABC 45.67% 700 ms
SVM 97.67% 600 ms

Figure 5. Overall time consumption of heuristic algorithms

9. CONCLUSION
In this paper, we analyze the accuracy of prediction of heart disease using machine learning
algorithm classifiers taking a dataset from UCI. The different classifiers of machine learning algorithms are
exploited for training and testing purposes. The main classifiers used are, support vector machine, bat, ABC,
and memory-based learner based on these results has achieved an accuracy of SVM 97.90% with 90.96%
sensitivity and 98.83% specificity the most accurate classifier among all of them. Future research directions

Int J Adv Appl Sci, Vol. 11, No. 2, June 2022: 158-167
Int J Adv Appl Sci ISSN: 2252-8814  165

include weighted voting-based classifier ensemble and application of the proposed algorithm on different
diseases like diabetes and cancer for classification and prediction.

REFERENCE
[1] L. Hausmann, F. Wirth, M. O. Flammer, J. Hofmann, and J. Fleischer, “Aligning vocational training to the electromobile
transformation by establishing the ‘training factory stator production’ - A methodical deficit analysis with derivation of
measures,” Procedia Manufacturing, vol. 45, pp. 448–453, 2020, doi: 10.1016/j.promfg.2020.04.051.
[2] A. R. M. Forkan, I. Khalil, Z. Tari, S. Foufou, and A. Bouras, “A context-aware approach for long-term behavioural change
detection and abnormality prediction in ambient assisted living,” Pattern Recognition, vol. 48, no. 3, pp. 628–641, 2015, doi:
10.1016/j.patcog.2014.07.007.
[3] A. J. Dobson, H. M. Alexander, R. F. Heller, and D. M. Lloyd, “How soon after quitting smoking does risk of heart attack
decline?,” Journal of Clinical Epidemiology, vol. 44, no. 11, pp. 1247–1253, 1991, doi: 10.1016/0895-4356(91)90157-5.
[4] R. B. D’Agostino, S. Grundy, L. M. Sullivan, and P. Wilson, “Validation of the Framingham coronary heart disease prediction
scores: Results of a multiple ethnic groups investigation,” Journal of the American Medical Association, vol. 286, no. 2, pp.
180–187, 2001, doi: 10.1001/jama.286.2.180.
[5] A. H. Chen, S. Y. Huang, P. S. Hong, C. H. Cheng, and E. J. Lin, “HDPS: Heart disease prediction system,” Computing in
Cardiology, vol. 38, pp. 557–560, 2011.
[6] S. H. Jee et al., “A coronary heart disease prediction model: The Korean heart study,” BMJ Open, vol. 4, no. 5, 2014, doi:
10.1136/bmjopen-2014-005025.
[7] K. F. Ng, M. A. Elwaifa, S. K. Paul, H. T. Lu, and C. Y. Lee, “A Rare Case of Ruptured Sinus of Valsalva Aneurysm in a
Young Patient Presenting with Heart Failure,” Journal of Cardiovascular Disease Research, vol. 11, no. 4, pp. 1–3, 2020, doi:
10.31838/jcdr.2020.11.04.01.
[8] S. A. Khan et al., “Brief tale of syncopation in peripartum cardiomyopathy unveiled as complete heart block -report of two
cases with review of literature,” Journal of Cardiovascular Disease Research, vol. 11, no. 3, pp. 90–94, 2020, doi:
10.31838/jcdr.2020.11.03.21.
[9] R. K. Abd, S. N. Abd, and V. Raman, “Tracing the risk factors of heart diseases at al-Nasiriyah heart center in Iraq,” Journal
of Cardiovascular Disease Research, vol. 10, no. 1, pp. 31–34, 2019, doi: 10.5530/jcdr.2019.1.6.
[10] D. Hemalatha and S. Poorani, “Machine learning techniques for heart disease prediction,” Journal of Cardiovascular Disease
Research, vol. 12, no. 1, pp. 93–96, 2021, doi: 10.31838/jcdr.2021.12.01.05.
[11] I. G. A. M. megayanti Ni Nyoman Wahyu Udayani, I Made Agus Sunadi Putra, Fitria Megawati, “The Rationality Use of
Diuretic Drug with Heart Failure to Inpatient Care at One of The Hospitals In Denpasar City,” Journal of Cardiovascular
Disease Research, vol. 12, no. 3, pp. 60–64, 2021, doi: 10.31838/jcdr.2021.12.03.08.
[12] S. S. Bunova, O. V. Zamahina, N. A. Nikolaev, N. I. Zhernakova, and A. A. Grishchenko, “The Prediction of the Bisoprolol
Effectiveness in Patients with Stable Coronary Artery Disease with Post-Infarction Cardiosclerosis,” Journal of
Cardiovascular Disease Research, vol. 11, no. 4, pp. 105–109, 2020, doi: 10.31838/jcdr.2020.11.04.18.
[13] D. L. Ramatillah, S. A. S. Sulaiman, A. H. Khan, D. O. L. Meng, and Markum, “Probability of dying and survival analysis of
diabetic and/hypertensive patients who undergone hemodialysis with heart disease complication: A comparison between a
hemodialysis center jakarta, Indonesia and penang, malaysia,” Journal of Cardiovascular Disease Research, vol. 9, no. 1, pp.
24–27, 2018, doi: 10.5530/jcdr.2018.1.6.
[14] O. A. Osipova, A. I. Golovin, R. Y. Shevtsov, N. I. Zhernakova, K. A. Bocharova, and Z. Y. Chefranova, “Ischemic heart
disease and risk of development of cognitive disorders or dementia (review),” Journal of Cardiovascular Disease Research,
vol. 11, no. 4, pp. 87–90, 2020, doi: 10.31838/jcdr.2020.11.04.15.
[15] G. Shanoli, R. Sanchita, K. Soumitra, P. Pritha, D. Atreyee, and H. Ajanta, “Homocysteine-Is there any role in Coronary
Heart Disease?,” Journal of Cardiovascular Disease Research, vol. 8, no. 2, pp. 46–49, 2017, doi: 10.5530/jcdr.2017.2.11.
[16] M. D. Salawu et al., “A chi-square-SVM based pedagogical rule extraction method for microarray data analysis,”
International Journal of Advances in Applied Sciences, vol. 9, no. 2, p. 93, 2020, doi: 10.11591/ijaas.v9.i2.pp93-100.
[17] S. N. Khan, N. M. Nawi, M. Imrona, A. Shahzad, A. Ullah, and Atta-ur-Rahman, “Opinion mining summarization and
automation process: A survey,” International Journal on Advanced Science, Engineering and Information Technology, vol. 8,
no. 5, pp. 1836–1844, 2018, doi: 10.18517/ijaseit.8.5.5002.
[18] S. N. Khan et al., “Comparative analysis for heart disease prediction,” International Journal on Informatics Visualization, vol.
1, no. 4–2, pp. 227–231, 2017, doi: 10.30630/joiv.1.4-2.66.
[19] A. Ullah, A., Şahin, C. B., Dinler, O. B., Khan, M. H., “Heart Disease Prediction Using Various Machines Learning
Approach,” Journal of Cardiovascular Disease Research, vol. 12, no. 3, pp. 379–391, 2021.
[20] H. Aznaoui, A. Ullah, S. Raghay, L. Aziz, and M. H. Khan, “An efficient GAF routing protocol using an optimized weighted
sum model in WSN,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 22, no. 1, pp. 396–406, 2021,
doi: 10.11591/ijeecs.v22.i1.pp396-406.
[21] A. Ullah, A. Salam, H. El Raoui, D. Sebai, and M. Rafie, “Towards more accurate iris recognition system by using hybrid
approach for feature extraction along with classifier,” International Journal of Reconfigurable and Embedded Systems, vol.
11, no. 1, pp. 59–70, 2022, doi: 10.11591/ijres.v11.i1.pp59-70.
[22] T. Alam, A. Ullah, and M. Benaida, “Deep reinforcement learning approach for computation offloading in blockchain -enabled
communications systems,” Journal of Ambient Intelligence and Humanized Computing, 2022, doi: 10.1007/s12652-021-
03663-2.
[23] A. Ullah and A. Chakir, “Improvement for tasks allocation system in VM for cloud datacenter using modified bat algorithm,”
Multimedia Tools and Applications, 2022, doi: 10.1007/s11042-022-12904-1.
[24] S. Ouhame, Y. Hadi, and A. Ullah, “An efficient forecasting approach for resource utilization in cloud data center using CNN -
LSTM model,” Neural Computing and Applications, vol. 33, no. 16, pp. 10043–10055, Aug. 2021, doi: 10.1007/s00521-021-
05770-9.
[25] R. Katarya and S. K. Meena, “Machine Learning Techniques for Heart Disease Prediction: A Comparative Study and
Analysis,” Health and Technology, vol. 11, no. 1, pp. 87–97, 2021, doi: 10.1007/s12553-020-00505-7.
[26] O. S. E. and A. A. E., “Effects of autoclaving (Pleurotus osteratus) spent mushroom substrate water extract on the minerals,
elicitors, cassava yield and the management of African cassava mosaic virus,” International Journal of Advances in Applied
Sciences, vol. 10, no. 1, p. 13, 2021, doi: 10.11591/ijaas.v10.i1.pp13-19.

Heart disease classification using various heuristic algorithms (Arif Ullah)

166  ISSN: 2252-8814

[27] A. Ullah, Ö. Batur Dinler, and C. Batur Şahin, “The Effect of Technology and Service on Learning Systems During the
COVID-19 Pandemic,” European Journal of Science and Technology, 2021, doi: 10.31590/ejosat.990073.
[28] A. Ullah and N. M. Nawi, “An improved in tasks allocation system for virtual machines in cloud computing using HBAC
algorithm,” Journal of Ambient Intelligence and Humanized Computing, 2021, doi: 10.1007/s12652-021-03496-z.
[29] A. Srinivasulu and A. Pushpa, “Disease prediction in big data healthcare using extended convolutional neural network
techniques,” International Journal of Advances in Applied Sciences, vol. 9, no. 2, p. 85, 2020, doi: 10.11591/ijaas.v9.i2.pp85-
92.
[30] A. Ullah, N. M. Nawi, J. Uddin, S. Baseer, and A. H. Rashed, “Artificial bee colony algorithm used for load balancing in
cloud computing: Review,” IAES International Journal of Artificial Intelligence, vol. 8, no. 2, pp. 156–167, 2019, doi:
10.11591/ijai.v8.i2.pp156-167.
[31] S. S. Tadasare and S. S. Pawar, “Early Detection of High Blood Pressure and Diabetic Retinopathy on Retinal Fundus Images
Using CBRIR Based on Lifting Wavelets,” International Journal of Advances in Applied Sciences, vol. 7, no. 4, p. 334, 2018,
doi: 10.11591/ijaas.v7.i4.pp334-346.
[32] A. T. Azar and S. A. El-Said, “Performance analysis of support vector machines classifiers in breast cancer mammography
recognition,” Neural Computing and Applications, vol. 24, no. 5, pp. 1163–1177, 2014, doi: 10.1007/s00521-012-1324-4.
[33] J. B. O. Mitchell B.O., “Machine learning methods in chemoinformatics,” Wiley Interdisciplinary Reviews: Computational
Molecular Science, vol. 4, no. 5, pp. 468–481, 2014, doi: 10.1002/wcms.1183.
[34] K. Armstrong, “Including electromagnetic interference (EMI) in functional safety risk assessments,” Achieving Systems Safety
- Proceedings of the 20th Safety-Critical Systems Symposium, SSS 2012, pp. 97–114, 2012, doi: 10.1007/978-1-4471-2494-
8_9.
[35] N. F. Matsatsinis and Y. Siskos, “Intelligent Decision Support Systems in Marketing,” City University London, pp. 273–336,
2003, doi: 10.1007/978-1-4615-1147-2_8.
[36] L. Parthiban and R. Subramanian, “Intelligent Heart Disease Prediction System using CANFIS and Genet ic Algorithm,”
International Journal of Biological and Medical Sciences, vol. 3, no. 3, pp. 157–160, 2008, doi: 10.5281/zenodo.1082439.
[37] M. Svensson, “Do not go breaking your heart: Do economic upturns really increase heart attack mortality?,” Social Science
and Medicine, vol. 65, no. 4, pp. 833–841, 2007, doi: 10.1016/j.socscimed.2007.04.015.
[38] M. Hendryx and K. J. Zullig, “Higher coronary heart disease and heart attack morbidity in Appalachian coal mining regions,”
Preventive Medicine, vol. 49, no. 5, pp. 355–359, 2009, doi: 10.1016/j.ypmed.2009.09.011.
[39] W. G. Rothstein, “Making Sense of Illness: Science, Society, and Disease,” Bulletin of the History of Medicine, vol. 73, no. 1,
pp. 181–182, 1999, doi: 10.1353/bhm.1999.0005.
[40] W. P. Cannell, “Height‐keeping reliability of aircraft at high altitude,” Quality and Reliability Engineering International, vol.
3, no. 4, pp. 273–280, 1987, doi: 10.1002/qre.4680030410.
[41] K. T. Khaw and E. Barrett-Connor, “Family history of heart attack: A modifiable risk factor,” Circulation, vol. 74, no. 2, pp.
239–244, 1986, doi: 10.1161/01.CIR.74.2.239.
[42] A. U. Sebai, D., & Shah, “Semantic-oriented learning-based image compression by Only-Train-Once quantized
autoencoders,” Signal, Image and Video Processing, pp. 1–9, 2022, doi: 10.1007/s11760-022-02231-1.

BIOGRAPHIES OF AUTHORS

Arif Ullah He is completed my Ph.D. in cloud computing with 2 years of

experience in teaching and research. Area of expertise in cloud computing, iot. Areas of
interest include software defined networking (SDN), load balancing, switches migration,
WSN, E-Learning, AI, WSN, and security. Email: [email protected]
[email protected].

Shakeel Ahmad Khan He is currently working as Assistant Professor in

Government College for Boys Chitral. He is doing is PhD in SBBU Pakistan. Areas of interest
include software defined networking (SDN), load balancing. Email:
[email protected].

Int J Adv Appl Sci, Vol. 11, No. 2, June 2022: 158-167
Int J Adv Appl Sci ISSN: 2252-8814  167

Tanweer Alam Associate Professor Faculty of Computer and Information

Systems, Islamic University of Madinah, Saudi Arabia Areas of interest include software
defined networking (SDN), load balancing, switches migration, WSN, E-Learning, AI, WSN,
& security. He can be contacted at email: [email protected].

Shkurte Luma-Osmani PhD Candidate in Computer Sciences Research

Assistant at University of Tetova North Macedonia. Areas of interest include software defined
networking (SDN), load balancing, switches migration, WSN, E-Learning, AI, WSN, &
security. Email: [email protected], [email protected].

Mahanz Sadie She is received the B.Sc. Degree in Computer Engineering

fromAllameh Mohaddese Noori Institute of Higher Education, Iran, in 2006, the M.Sc. Degree
inComputer Architecture from Islamic Azad University of Arak in 2011, the Ph.D. Degree
inComputer Architecture at Department of Computer Engineering, Science and
ResearchBranch, Islamic Azad University, Tehran, Iran, in 2016. She can be contacted at
email: [email protected].

Heart disease classification using various heuristic algorithms (Arif Ullah)