Table 8. Comparison of results achieved on the MIT-BIH dataset.

Reference	Methods	Accuracy (%)
Acharya et al. (2017)	Custom 9-layer deep CNN	94.03
Ullah et al. (2020)	2D CNN model applied on FFT spectrograms	99.11
Jin et al. (2020)	Domain Adaptive Residual Network
Li et al. (2020)	Fully-connected neural networks as classifiers on handcraft features	89.25
Romdhane et al. (2020)	Custom 1D CNN	98.41
Van Steenkiste, Van Loon & Crevecoeur (2020)	Custom CNN optimized by Genetic Algorithm (GA)	97.7
Yang et al. (2021)	Ensemble of mixed-kernel extreme learning machine-based random forest binary classifiers	98.1
This article	Deep features from 3 deep networks and Cubic SVM	97.6