Table 9.

Prediction results of PSO-LSSVM based on hyperspectral data .

Feature extraction method	Number of feature variables	${\text{R}}_c^2$	RMSEC	${\mathit{\text{R}}}_p^2$	RMSEP	RPD
Boss	19	0.8403	0.7085	0.8056	0.6227	2.02
CARS	35	0.8126	0.7676	0.7710	0.6758	2.03
IVSO	44	0.8231	0.7456	0.8532	0.5411	2.54
IVISSA	70	0.8964	0.5710	0.7576	0.6953	1.94
MASS	53	0.8817	0.6098	0.7698	0.6775	1.93
CARS-Boss	27	0.8273	0.7367	0.7805	0.6616	2.10
MASS-Boss	21	0.8265	0.7385	0.8169	0.6042	2.28
IVISSA-Boss	17	0.8651	0.6512	0.7435	0.7152	2.01

The bold values represent the best performer in each table.