A Novel Discrete DGM (2, 1, ) Model Based on Stochastic Oscillation Sequence and Its Application

Table 1

Comparison of modeling accuracy of four models based on four stochastic oscillation sequences with different amplitudes.

Sequence data (2005–2010)	Model	Simulation value	Mean relative error (%)	Prediction value (2011)	Mean prediction error (%)
	1	0.874, 1.874, 1.262, 2.009, 1.400, 2.152	11.558	1.547	19.232
	2	0.874, 1.874, 1.390, 1.679, 1.770, 1.800	0.000	1.914	0.001
	3	0.874, 1.874, 1.802, 1.382, 1.650, 1.648	10.416	2.429	26.817
	4	0.874, 1.874, 1.390, 1.679, 1.770, 1.800	0.000	1.915	0.000
	1	0.874, 1.278, 2.257, 1.395, 2.376, 1.515	27.319	2.498	30.471
	2	0.874, 1.278, 1.390, 2.874, 1.770, 1.800	0.000	1.914	0.001
	3	0.874, 1.278, 3.688, 0.015, 3.109, 0.289	70.738	4.451	132.423
	4	0.874, 1.278, 1.390, 2.874, 1.770, 1.800	0.000	1.915	0.000
	1	0.874, 1.278, 4.033, 0.741, 3.521, 0.523	40.806	3.057	59.644
	2	0.874, 1.278, 3.874, 1.679, 1.770, 1.800	0.000	1.914	0.001
	3	0.874, 1.278, 5.446, −0.919, 4.168, −1.055	81.582	5.451	184.621
	4	0.874, 1.278, 3.874, 1.679, 1.770, 1.800	0.000	1.915	0.000
	1	0.874, 1.278, 2.323, 1.731, 2.794, 2.221	22.771	3.302	72.439
	2	0.874, 1.278, 1.390, 1.679, 4.874, 1.800	0.000	1.916	0.001
	3	0.874, 1.278, 5.788, −1.364, 4.652, −0.572	105.662	7.914	313.310
	4	0.874, 1.278, 1.390, 1.679, 4.874, 1.800	0.000	1.915	0.000