Arrhythmia Classification Techniques Using Deep Neural Network

Table 2

Arrhythmia classification techniques from 2010 to 2020.

Ref.	Publication year	No. of leads	Classification techniques	Optimization techniques	Accuracy (%)
[9]	2020	02	CNN	BAROA	93.19
[10]	2020	12	Extreme gradient boosting tree	Low pass filter	97
[11]	2020	02	1D-CNN	—	95
[12]	2019	04	DNN	PCA	97.8
[13]	2019	03	DNN	—	92.07
[14]	2019	05	1 D CNN-2D CNN	—	90.93
[15]	2019	05	Residual networks	Data augmentation	99.81
[16]	2018	06	EMD	LDA	87
[17]	2018	05	CNN LSTM	DL	98.10
[18]	2018	02	Deep belief networks	—	95.57
[19]	2017	02	DNN	—	92
[20]	2017	04	SVM	—	98.9
[21]	2017	05	GRNN	—	88
[22]	2016	05	NN	—	97
[23]	2016	02	Dynamic Bayesian	PCA	99
[24]	2015	03	ANFIS	—	96
[25]	2014	03	Feed forward PNN	—	96.5
[26]	2014	16	SVM	PCA	86
[27]	2014	5	ML classifier	—	99.48
[28]	2013	05	SVM	PCA-LDA	99.28
[29]	2013	05	NN	PCA	94.52
[30]	2013	02	MLPNN	—	95.1
[31]	2013	02	MLPNN	—	85
[32]	2013	03	BMLPNN	—	76
[33]	2012	02	MNN-generalized FFNN	—	86.67
[34]	2012	08	PNN	PCA-LDA	99.71
[35]	2011	05	NN	—	95
[36]	2011	02	FCM	—	99.05
[37]	2011	03	MLPNN	—	96.7
[38]	2010	05	MDPSO	—	95.58