A Machine Learning-Based Big EEG Data Artifact Detection and Wavelet-Based Removal: An Empirical Approach

Table 3

Comparison of EEG artifact removal performance before and after optimization.

Methodology	Evaluation parameters	Signal-to-noise ratio (SNR)
		10	20	25
EEMD-CCA-SWT	DSNR	27.2775	32.8229	31.3270
EEMD-CCA-SWT + HHO		28.2345	34.2475	37.2172
EEMD-CCA-SWT	Lambda	75.6307	88.746	87.2213
EEMD-CCA-SWT + HHO		78.6307	89.627	89.6127
EEMD-CCA-SWT	Correlation improvement	0.0161	0.0258	0.0140
EEMD-CCA-SWT + HHO		0.0141	0.0261	0.0121
EEMD-CCA-SWT	Spectral distortion (Pdis)	0.9640	0.9856	0.9867
EEMD-CCA-SWT + HHO		0.9356	0.9756	0.952
EEMD-CCA-SWT	RMSE	0.093	0.1126	0.0974
EEMD-CCA-SWT + HHO		0.091	0.1123	0.0913
EEMD-CCA-SWT	Coherence improvement () in percentage	86.93	84.26	85.11
EEMD-CCA-SWT + HHO		85.82	84.22	83.09