Development of Multidecomposition Hybrid Model for Hydrological Time Series Analysis

Table 3

Evaluation index of testing prediction error of proposed models (EMD-CEEMDAN-MM and WA-CEEMDAN-MM) with all selected models for all four case studies.

River Inflow	Model Name	Models	MRE	MAE	MSE
Indus Inflow	1-S	ARIMA	4.2347	0.0685	64.7141
	2-S	WA-ARMA	3.2862	0.0430	53.4782
		WA-RGMDH	3.2548	0.0393	46.7382
		WA-RBFN	20.1949	0.2598	2301.772
		EMD-ARMA	4.9898	0.0960	76.1440
		EMD-RGMDH	4.9653	0.0915	76.0884
		EMD-RBFN	34.3741	0.7762	3931.601
	3-S	EMD-EEMD-MM	5.2710	0.1721	44.0115
		WA-CEEMDAN-MM	1.5410	0.0349	5.5734
		EMD-CEEMDAN-MM	1.8009	0.0462	6.6983
Jhelum Inflow	1-S	ARMA	3.5224	0.1201	47.5529
	2-S	WA-ARMA	2.6129	0.0748	37.1441
		WA-RGMDH	2.6208	0.0773	37.7954
		WA-RBFN	9.8608	0.7714	180.7443
		EMD-ARMA	3.7354	0.1551	48.3164
		EMD-RGMDH	3.7357	0.1620	48.3606
		EMD-RBFN	2.8822	0.2506	51.9916
	3-S	EMD-EEMD-MM	2.0096	0.1269	7.3565
		WA-CEEMDAN-MM	1.1805	0.0457	6.8225
		EMD-CEEMDAN-MM	1.4480	0.0642	7.7709
Kabul Inflow	1-S	ARMA	2.4910	0.0883	25.0136
	2-S	WA-ARMA	1.9999	0.0592	20.6874
		WA-RGMDH	2.0794	0.0729	21.0612
		WA-RBFN	1.6565	0.0997	13.3554
		EMD-ARMA	2.9538	0.1484	28.5767
		EMD-RGMDH	3.0114	0.2280	28.9351
		EMD-RBFN	4.9355	0.7613	69.9346
	3-S	EMD-EEMD-MM	1.8758	0.3166	5.8020
		WA-CEEMDAN-MM	0.7664	0.0363	2.1072
		EMD-CEEMDAN-MM	0.9599	0.0861	2.7636
Chenab Inflow	1-S	ARMA	5.4157	0.4646	108.185
	2-S	WA-ARMA	3.9652	0.1087	84.2359
		WA-RGMDH	3.6147	0.0943	81.6493
		WA-RBFN	4.1424	0.2757	47.6184
		EMD-ARMA	4.7971	0.2721	100.7013
		EMD-RGMDHA	4.4812	0.1865	95.6680
		EMD-RBFN	10.8228	2.1666	284.5627
	3-S	EMD-EEMD-MM	2.7172	0.2298	14.5191
		WA-CEEMDAN-MM	1.6940	0.0705	13.5702
		EMD-CEEMDAN-MM	1.9345	0.1105	14.067