Improved GSO Optimized ESN Soft-Sensor Model of Flotation Process Based on Multisource Heterogeneous Information Fusion

Table 4

Predictive error comparison of soft-sensor model.

Predicted variables	Predictive method	NRMSE	MSE	MAPE
Concentrate grade (%)	ESN	0.0122	0.6828	1.0310
	GA-ESN	0.0093	0.2379	0.6020
	PSO-ESN	0.0089	0.1918	0.5435
	GSO-ESN	0.0102	0.3562	0.7205
	IGSO-ESN	0.0069	0.0681	0.3306
Flotation recovery rate (%)	ESN	0.0096	1.0195	0.8807
	GA-ESN	0.0075	0.3593	0.5468
	PSO-ESN	0.0078	0.4158	0.5929
	GSO-ESN	0.0088	0.6626	0.7482
	IGSO-ESN	0.0060	0.1467	0.3530