Short-Term Prediction of Air Pollution in Macau Using Support Vector Machines

Table 6

The tested results of SPM, SO₂, NO₂, and O₃ in different kernel models in summer experiment.

	Kernel
	Linear	Polynomial	RBF	Sigmoid	Wavelet
SPM
MAE	0.165	0.233	0.169	0.438	0.258
RMSE	0.243	0.280	0.248	0.685	0.379
CWIA	34.8%	48.1%	34.1%	62.2%	41.5%
RE	29.0%	54.8%	32.3%	58.1%	45.2%
SO2
MAE	0.069	0.221	0.058	0.748	0.111
RMSE	0.089	0.244	0.086	1.022	0.164
CWIA	9.3%	43.9%	8.8%	68.6%	24.1%
RE	48.4%	93.5%	32.3%	100.0%	61.3%
NO2
MAE	0.653	0.719	0.675	3.646	0.578
RMSE	0.809	1.352	0.824	9.418	0.723
CWIA	76.7%	92.4%	77.1%	96.6%	78.6%
RE	67.7%	41.9%	67.7%	80.6%	64.5%
O3
MAE	0.165	0.154	0.170	1.495	0.322
RMSE	0.227	0.237	0.243	1.932	0.435
CWIA	25.5%	30.3%	24.6%	77.5%	47.0%
RE	29.0%	25.8%	22.6%	90.3%	61.3%