An Efficient Machine Learning Model Based on Improved Features Selections for Early and Accurate Heart Disease Predication

Table 1

Previous literature review.

Reference	Heart disease type	Application	ML algorithm	Approach	Evaluations (%)	Data
[23]	Coronary disease	Classification	CA, BA	Undersampling	71.1	425 patients data
[24]	General heart disease	Classification	MLP	Undersampling	80	Cleveland dataset
[25]	General heart disease	Classification	ANN	Sampling	84	Cleveland dataset
[26]	General heart disease	Three-phase system for the prediction	ANN	Data sampling	85	Uci
[5]	Heart disease	Ensemble-based predictive model	ANN	Undersampling	91	Cleveland heart disease
[27]	Coronary Heart disease	Adaptive fuzzy ensemble	GA, MS-pso	Feature selection	92.31	Public dataset
[28]	Coronary artery disease	Classification	SVM, NB	Feature selection	96	Z-Alizadeh sani dataset
[29]	Cardiac disease	Classification	SVM, DT, KNN, etc.	Focal loss	86	Cleveland heart disease
[30]	Cardiac arrest	Scoring system classification	SVM	Undersampling	78.8	1386 records
[31]	Heart disease (general)	Detection	NB, SMO	Features selection	83	Cleveland dataset
[32]	Coronary heart disease	Predication	SVM, KNN, etc.	SMOTE	72	African heart disease data
[33]	Arrhythmia	Diagnosing	SVM, KNN, DT, RF	SMOTE	92	MIT-BIH
[34]	Heart arrhythmia	Detection	XGBoost classifier	Undersampling	87	Biobank UK dataset
[35]	Chronic heart failure (HF)	Incremental and boosting features value	DT, RF, SVM, KNN, LMT	Undersampling	89	487 patient data
[36]	Cardiovascular diseases	Classification	RF, DT	SMOTE	91	4270 patients data
[37]	Heart disease (general)	Features method	Lda, KNN, SVM, RF	Sampling	84	UCI dataset
[38]	Heart arrhythmia	Classification	Marine predators algorithm, SGD, CNN	Sampling	99.47	MIT-BIH arrhythmia, European, INCART
[39]	Heart arrhythmia	Classification	Marine predators algorithm, DNN, CNN	Sampling	99	MIT-BIH, EDB, and INCART
[38]	Heart arrhythmia	Classification	Manta ray foraging optimization, SVM, LBP, HOS	Sampling	98.26	MIT-BIH arrhythmia