A Hybrid Approach by Integrating Brain Storm Optimization Algorithm with Grey Neural Network for Stock Index Forecasting

Table 3

Fitting error results of different models under different data preprocessing operations.

Data preprocessing	Error type	Model	Training stage
			SSE Composite Index	Shenzhen Composite Index	HuShen 300 Index
Without normalization	MAE	GNN	26.0267	14.2249	36.0929
		BSO-GNN	6.4482	5.3152	10.8070
	RMSE	GNN	34.4066	18.5165	49.4477
		BSO-GNN	9.0112	6.8256	14.8538
	MAPE	GNN	1.1191	1.5722	1.4410
		BSO-GNN	0.2740	0.5813	0.4247
With normalization	MAE	GNN	0.0822	0.0701	0.0859
		BSO-GNN	0.0204	0.0262	0.0257
	RMSE	GNN	0.1086	0.0912	0.1177
		BSO-GNN	0.0285	0.0336	0.0354
	MAPE	GNN	19.2278^b	28.3084^b	20.6714^b
		BSO-GNN	4.0525^b	9.4707^b	5.2068^b

bCorrected value.