Table 5.

Cotton yield estimation results based on vegetation indices and texture features.

Feature selection	Data set	PLSR			Elastic-Net			KRR			SVR			MLP			ELM
		R 2	RMSE (t.ha−1)	rRMSE (%)	R 2	RMSE (t.ha−1)	rRMSE (%)	R 2	RMSE (t.ha−1)	rRMSE (%)	R 2	RMSE (t.ha−1)	rRMSE (%)	R 2	RMSE (t.ha−1)	rRMSE (%)	R 2	RMSE (t.ha−1)	rRMSE (%)
P	Cal	0.47	1.9824	70.42	0.47	1.9971	70.95	0.55	1.7202	64.66	0.51	1.9044	67.65	0.37	2.1631	76.84	0.75	1.4971	48.48
	Val	0.38	2.4432	70.39	0.38	2.4650	71.02	0.45	2.3235	66.94	0.42	2.3757	68.44	0.39	2.4377	70.23	0.70	1.6364	48.61
MIC	Cal	0.49	1.9446	69.66	0.49	1.9608	69.66	0.59	1.7382	61.75	0.53	1.8661	66.29	0.54	1.8334	65.13	0.78	1.3561	42.63
	Val	0.39	2.4048	69.28	0.40	2.4267	69.91	0.49	2.2275	64.17	0.44	2.3258	67.01	0.45	2.3142	66.67	0.75	1.5028	47.06
RF	Cal	0.44	2.0358	72.32	0.43	2.0754	73.73	0.73	1.4283	50.74	0.63	1.6923	60.12	0.65	1.6184	57.49	0.91	0.9277	29.34
	Val	0.33	2.5277	72.82	0.36	2.5190	72.57	0.62	1.9348	55.74	0.52	2.2403	64.54	0.49	2.2237	64.07	0.80	1.1699	37.64
RFE	Cal	0.47	1.9738	70.05	0.44	2.0503	72.76	0.53	1.8581	65.94	0.48	1.9775	70.18	0.46	1.9960	70.83	0.93	0.7279	25.88
	Val	0.38	2.4559	70.87	0.35	2.5136	72.54	0.45	2.3231	67.04	0.41	2.4028	69.34	0.31	2.5709	74.19	0.91	0.8516	31.64

The values in bold represent models with the best results in linear and nonlinear methods.