TY  - JOUR
EP  - 266
SN  - 21945357
PB  - Springer Verlag
TI  - Feature selection method based on grey wolf optimization for coronary artery disease classification
SP  - 257
N1  - cited By 56; Conference of 3rd International Conference of Reliable Information and Communication Technology, IRICT 2018 ; Conference Date: 23 June 2018 Through 24 June 2018; Conference Code:218299
AV  - none
VL  - 843
UR  - https://www.scopus.com/inward/record.uri?eid=2-s2.0-85053898760&doi=10.1007%2f978-3-319-99007-1_25&partnerID=40&md5=7e695b86906d292d455039979f6795d5
JF  - Advances in Intelligent Systems and Computing
A1  - Al-Tashi, Q.
A1  - Rais, H.
A1  - Jadid, S.
Y1  - 2019///
KW  - Cardiology; Diagnosis; Diseases; Feature extraction; Heart; Soft computing; Support vector machines
KW  -  Cardio-vascular disease; Classification accuracy; Coronary artery disease; Feature selection methods; Feature selection techniques; Fitness functions; Performance validation; Support vector machine classifiers
KW  -  Classification (of information)
ID  - scholars12305
N2  - Cardiovascular disease has been declared as one of the deadly illness that affects humans in the Middle and Old ages across the globe. One of the cardiovascular disease known as Coronary artery, has recorded the highest number of motility rates in the recent years. Machine learning tools have been very effective in investigating the causes of such lethal disease which involve analyzing large amount of dataset. Such datasets might contain redundant and irrelevant features which affect the classification accuracy and processing speed. Hence, applying feature selection technique for the elimination of the said redundant and irrelevant features is necessary. In this paper, a novel wrapper feature selection method is proposed to determine the optimal feature subset for diagnosing coronary artery disease. This proposed method consists of two main stages feature selection and classification. In the first stage, Grey Wolf Optimization (GWO) is used to find the best features in the disease identification dataset. In the second stage, the fitness function of GWO is evaluated using Support Vector Machine classifier (SVM). Cleveland Heart disease dataset is used for performance validation of the proposed method. The experimental results showed that, the proposed GWO-SVM outperforms current existing approaches with an achievement of 89.83 in accuracy, 93 in sensitivity and 91 in specificity rates. © Springer Nature Switzerland AG 2019.
ER  -