Table 2.

Performance criteria of DTs for plant type identification

DT	Feature selection/reduction method	Train			Test
		Kappa	RMSE	Accuracy (%)	Kappa	RMSE	Accuracy (%)
J48	–a	0.70	0.3129	77.50	0.64	0.3483	73.61
	CFS	0.74	0.2855	80.83	0.73	0.2947	80.56
	PCA	0.39	0.4594	54.17	0.36	0.4859	52.77
REP	–a	0.70	0.3135	77.50	0.60	0.3334	75.00
	CFS	0.73	0.3007	80.00	0.72	0.3143	79.17
	PCA	0.38	0.4121	53.33	0.34	0.4930	51.38
RT	–a	0.62	0.3764	71.67	0.57	0.3953	68.75
	CFS	0.72	0.3227	79.17	0.70	0.3371	77.27
	PCA	0.40	0.4743	55.00	0.38	0.4556	54.17

Bolditalic value indicates the most accurate DT classifier

^aNo feature selection applied (classification using all input features)