Comparison of Three Prediction Models for Predicting Chronic Obstructive Pulmonary Disease in China

Figures & data

Table 1 Fitting and Prediction Effect Evaluation of the Three Models

Model	R^2	MAE	MAPE (%)	RMSE	AIC	BIC	Relative Error of Prediction (%)
GAM	0.9999	6929.9	0.02	9748.8	660.77	692.05	−0.0003
ARIMA (0,1,2)	0.9973	138,430.7	0.36	238,583.1	784.12	789.44	−0.4100
Curve Fitting Method	0.9866	425,550.3	1.13	521,219.8	853.81	859.28	−4.1891

Abbreviations: GAM, generalized additive model; ARIMA, autoregressive integrated moving average; MAE, mean absolute error; MAPE, mean absolute percentage error; RMSE, root mean squared error; AIC, Akaike information criterion; BIC, Bayesian information criterion.

Table 2 Fitting and Prediction Results of the Three Models from 1990 to 2025

Year	Actual Values	Fitted/Predicted* Values of GAM	Fitted/Predicted* Values of ARIMA (0,1,2)	Fitted/Predicted* Values of Curve Fitting Method
1990	27,750,213	27,749,770	27,723,071	28,151,525
1991	28,631,432	28,633,235	28,473,099	28,934,960
1992	29,479,714	29,476,282	29,431,996	29,699,679
1993	30,298,665	30,303,071	30,204,186	30,445,684
1994	31,130,857	31,126,443	31,052,313	31,172,974
1995	31,948,269	31,951,696	31,882,342	31,881,550
1996	32,759,336	32,758,004	32,678,443	32,571,411
1997	33,495,325	33,493,665	33,503,401	33,242,557
1998	34,189,045	34,193,573	34,128,844	33,894,989
1999	34,898,215	34,891,921	34,873,317	34,528,706
2000	35,562,685	35,569,183	35,566,508	35,143,708
2001	36,229,393	36,224,555	36,176,711	35,739,996
2002	36,812,918	36,814,608	36,905,493	36,317,569
2003	37,321,211	37,323,028	37,321,493	36,876,427
2004	37,800,883	37,796,257	37,886,149	37,416,571
2005	38,203,544	38,209,824	38,294,940	37,938,000
2006	38,503,279	38,497,066	38,650,508	38,440,714
2007	38,606,128	38,609,871	38,873,094	38,924,714
2008	38,624,314	38,624,160	38,790,938	39,389,999
2009	38,750,857	38,746,558	38,887,876	39,836,570
2010	39,090,394	39,100,023	39,099,805	40,264,426
2011	39,762,225	39,748,085	39,622,811	40,673,567
2012	40,571,705	40,589,640	40,551,258	41,063,993
2013	41,450,223	41,429,365	41,270,513	41,435,705
2014	42,372,472	42,396,174	42,306,588	41,788,703
2015	43,108,193	43,084,290	43,150,240	42,122,985
2016	42,698,723	42,713,748	43,690,108	42,438,553
2017	42,426,516	42,422,833	41,967,170	42,735,407
2018	43,608,453	43,608,259	43,191,680	43,013,545
2019	45,164,960	45,164,812	44,981,522	43,272,969
2020	—	46,721,365	45,780,166	43,513,679
2021	—	48,277,917	46,387,883	43,735,674
2022	—	49,834,470	46,995,600	43,938,954
2023	—	51,391,023	47,603,317	44,123,519
2024	—	52,947,576	48,211,033	44,289,370
2025	—	54,504,128	48,818,750	44,436,506

Note: *Predicted values: the numbers in bold are the predicted values.

Abbreviations: GAM, generalized additive model; ARIMA, autoregressive integrated moving average.

Figure 1 The prevalence trend of COPD in China from 1990 to 2025.

Abbreviations: COPD, chronic obstructive pulmonary disease; GAM, generalized additive model; ARIMA, autoregressive integrated moving average.