Forecasting the Agriculture Output Values in China Based on Grey Seasonal Model

Table 4

Fitting results of the GSM, Holt–Winters model, and SARIMA model (100 million yuan).

Year	Quarter	Actual value	GSM value	Holt–Winters value	SARIMA value	SDGM
2012	1	1249.7	1266.8	1249.7	1249.7
	2	2917.7	2932.4	2917.7	2917.7
	3	4857.0	4878.2	4857.0	4857.0
	4	8706.0	8653.0	8706.0	8706.0
2013	1	1390.2	1400.8	1390.2	1324.8	1390.20
	2	3220.6	3242.7	972.0	3226.4	3140.22
	3	5362.6	5394.2	5362.6	5344.8	5094.93
	4	9634.6	9568.5	9730.9	9416.4	8911.22
2014	1	1484.9	1495.9	1484.6	1562.0	1066.81
	2	3455.5	3462.9	3415.2	3471.7	3178.07
	3	5773.0	5760.6	5868.0	5801.3	5469.01
	4	10334.3	10218.4	10462.9	10354.5	9757.43
2015	1	1586.4	1597.6	1589.9	1610.2	1302.39
	2	3668.7	3698.2	3643.2	3733.6	3551.56
	3	6111.8	6151.9	6222.0	6199.0	6036.10
	4	10880.6	10912.5	11062.1	11012.5	10656.68
2016	1	1731.2	1706.1	1672.0	1665.7	1745.19
	2	3960.4	3949.4	3964.9	4007.5	4063.07
	3	6575.6	6569.8	6708.6	6575.7	6604.64
	4	11602.9	11653.8	11885.7	11555.3	11324.88
2017	1	1850.2	1822.0	1781.1	1847.4	1855.38
	2	4288.4	4217.6	4231.7	4292.8	4368.11
	3	7104.3	7016.1	7251.4	7051.4	7068.99
	4	12316.9	12445.4	12825.1	12295.9	12000.87
MAPE		0.75%	5.19%	1.31%	4.89%