A Hybrid Classification System for Heart Disease Diagnosis Based on the RFRS Method

Table 6

Comparison of our results with those of other studies.

Author	Method	Classification accuracy (%)
Our study	RFRS classification system	92.59
Lee [4]	Graphical characteristics of BSWFM combined with Euclidean distance	87.4
Tomar and Agarwal [5]	Feature selection-based LSTSVM	85.59
Buscema et al. [6]	TWIST algorithm	84.14
Subbulakshmi et al. [7]	ELM	87.5
Karegowda et al. [8]	GA + Naïve Bayes	85.87
Srinivas et al. [9]	Naïve Bayes	83.70
Polat and Güneş [10]	RBF kernel -score + LS-SVM	83.70
Özşen and Güneş [11]	GA-AWAIS	87.43
Helmy and Rasheed [12]	Algebraic Sigmoid	85.24
Wang et al. [13]	Linear kernel SVM classifiers	83.37
Özşen and Güneş [14]	Hybrid similarity measure	83.95
Kahramanli and Allahverdi [15]	Hybrid neural network method	86.8
Yan et al. [16]	ICA + SVM	83.75
Şahan et al. [17]	AWAIS	82.59
Duch et al. [18]	KNN classifier	85.6

BSWFM: bounded sum of weighted fuzzy membership functions; LSTSVM: Least Square Twin Support Vector Machine; TWIST: Training with Input Selection and Testing; ELM: Extreme Learning Machine; GA: genetic algorithm; SVM: support vector machine; ICA: imperialist competitive algorithm; AWAIS: attribute weighted artificial immune system; NN: -nearest neighbor.