Multiclass Classification of Cardiac Arrhythmia Using Improved Feature Selection and SVM Invariants

Table 4

Comparison of the proposed model with state-of-the-art methods.

Method	Train-test split	Number of selected features	Accuracy (%)
Proposed	90-10	94	92.07%
IBPLN + LM [16]	68-32	18	87.71%
Modular NN [15]	90-10	198	78.89%
SVM [18]	75-25	60	84%
NB [26]	70-30	205	70.50%
KNN [24]	20-fold CV	148	73.80%
MLP NN [13]	3-fold CV	-	88.24%
NN [14]	90-10	79	76.67%