A Feature-Driven Decision Support System for Heart Failure Prediction Based on Statistical Model and Gaussian Naive Bayes

Table 5

Details of other machine learning methods proposed for HF prediction and their obtained HF prediction accuracies.

Study (year)	Method	Accuracy (%)
ToolDiag, RA [26]	IB1-4	50.00
WEKA, RA [26]	InductH	58.50
ToolDiag, RA [26]	RBF	60.00
WEKA, RA [26]	FOIL	64.00
ToolDiag, RA [26]	MLP + BP	65.60
WEKA, RA [26]	T2	68.10
WEKA, RA [26]	1R	71.40
WEKA, RA [26]	IB1c	74.00
WEKA, RA [26]	K	76.70
Robert Detrano [26]	Logistic regression	77.00
Cheung (2001) [27]	C4.5	81.11
Cheung (2001) [27]	Naive Bayes	81.48
Cheung (2001) [27]	BNND	81.11
Cheung (2001) [27]	BNNF	80.96
WEKA, RA [26]	Naive Bayes	83.60
Ster and Dobnikar [28]	Fisher discriminant analysis	84.2
Ster and Dobnikar [28]	Linear discriminant analysis	84.5
Ster and Dobnikar [28]	Naive Bayes	82.5–83.4
Polat et al. (2005) [18]	AIRS	84.50
Ozsen et al. (2005) [29]	Kernel functions with AIS	85.93
Kahramanli and Allahverdi (2008) [30]	Hybrid neural network system	86.8
Polat et al. (2006) [1]	Fuzzy-AIRS-Knn-based system	87.00
Özşen and Güneş (2009) [19]	Modified artificial immune system	87.43
Das et al. (2009) [2]	Neural network ensembles	89.01
Jankowski and Kadirkamanathan (1997) [31]	IncNet	90.00
Kumar (2011) [32]	ANFIS	91.18
Samuel et al. (2017) [20]	ANN-fuzzy-AHP	91.10
Kumar (2012) [33]	Fuzzy resolution mechanism	91.83
Ali et al. (2019) [21]	Stacked and optimized SVMs	92.22
Paul et al. (2018) [23]	Adaptive weighted fuzzy system ensemble	92.31
Proposed method (2018)	-GNB	93.33