COVID-19 epidemic forecasting based on a comprehensive ensemble method

BAI Yun, QIAN Zhen, SUN Yuying, WANG Shouyang

Systems Engineering - Theory & Practice ›› 2022, Vol. 42 ›› Issue (6) : 1678-1693.

PDF(1188 KB)
中国科学院数学与系统科学研究院期刊网
PDF(1188 KB)
Systems Engineering - Theory & Practice ›› 2022, Vol. 42 ›› Issue (6) : 1678-1693. DOI: 10.12011/SETP2021-3005

COVID-19 epidemic forecasting based on a comprehensive ensemble method

  • BAI Yun1,2, QIAN Zhen1,2, SUN Yuying1,2,3, WANG Shouyang1,2,3
Author information +
History +

Abstract

Since December 2019, COVID-19 epidemic is continuing to spread globally. It not only jeopardizing the lives and health of people around the world seriously and putting a severe test on the public medical and health system, but also causes a huge impact on economic and trade activities and has a deep influence on the international community. In order to help researchers and policy makers understand the mechanism of virus transmission and adopt reasonable anti-epidemic policies to inhibit the further spread of the virus, some studies have adopted mathematical prediction models to simulate the spread of the virus and the development of the epidemic. However, the existing research has certain limitations, such as single method selection, excessive reliance on model parameters selection, and virus transmission and policy adjustments caused time variability of data. To solve the above problems, this paper proposes a comprehensive ensemble forecasting framework, which bases on six single prediction models, including time-varying Jackknife model averaging (TVJMA), time-varying parameters (TVP), time-varying parameter SIR (vSIR), logistic regression (LR), polynomial regression (PNR), autoregressive moving average (ARMA). The proposed method is used to predict the cumulative number of confirmed cases in the 6 most severely affected countries in different regions. Empirical results show that for a single prediction method, the TVJMA method outperforms the other five methods; the comprehensive ensemble forecasting method is significantly better than any single method in most cases, especially, the multi-model combined forecasting method based on error correction weights improves the prediction accuracy significantly. For different prediction steps, the comprehensive ensemble forecasting method is robust.

Key words

COVID-19 / ensemble forecasting method / time-varying Jackknife model averaging / non-parametric estimation / time-varying parameter SIR / MCS test

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations
BAI Yun , QIAN Zhen , SUN Yuying , WANG Shouyang. COVID-19 epidemic forecasting based on a comprehensive ensemble method. Systems Engineering - Theory & Practice, 2022, 42(6): 1678-1693 https://doi.org/10.12011/SETP2021-3005

References

[1] Rothe C, Schunk M, Sothmann P, et al. Transmission of 2019-nCOV infection from an asymptomatic contact in germany[J]. New England Journal of Medicine, 2020, 382(10):970-971.
[2] Li R, Pei S, Chen B, et al. Substantial undocumented infection facilitates the rapid dissemination of novel coronavirus (SARS-CoV-2)[J]. Science, 2020, 368(6490):489-493.
[3] Pal R, Sekh A A, Kar S, et al. Neural network based country wise risk prediction of COVID-19[J]. Applied Sciences, 2020, 10(18):6448.
[4] Achterberg M A, Prasse B, Ma L, et al. Comparing the accuracy of several network-based COVID-19 prediction algorithms[J]. International Journal of Forecasting, 2022, 38(2):489-504.
[5] 张发,李璐,宣慧玉.传染病传播模型综述[J].系统工程理论与实践, 2011, 31(9):1736-1744.Zhang F, Li L, Xuan H Y. Survey of transmission models of infectious diseases[J]. Systems Engineering-Theory&Practice, 2011, 31(9):1736-1744.
[6] 金启轩.中国新冠肺炎疫情预测建模与理性评估[J].统计与决策, 2020(2):1-8.Jin Q X. Novel coronavirus pneumonia epidemic prediction modeling and rational evaluation in China[J]. Statistics&Decision, 2020(2):1-8.
[7] 严阅,陈瑜,刘可#
[157],等.基于一类时滞动力学系统对新型冠状病毒肺炎疫情的建模和预测[J].中国科学:数学, 2020, 50(3):385-392.Yan Y, Chen Y, Liu K J, et al. Modeling and prediction for the trend of outbreak of NCP based on a time-delay dynamic system[J]. Scientia Sinica (Mathematica), 2020, 50(3):385-392.
[8] Chen Y, Cheng J, Jiang Y, et al. A time delay dynamical model for outbreak of 2019-nCoV and the parameter identification[J]. Journal of Inverse and Ill-Posed Problems, 2020, 28(2):243-50.
[9] 桑茂盛,丁一,包铭磊,等.基于新冠病毒特征及防控措施的传播动力学模型[J].系统工程理论与实践, 2021, 41(1):124-133.Sang M S, Ding Y, Bao M L, et al. Propagation dynamics model considering the characteristics of 2019-nCoV and prevention-control measurements[J]. Systems Engineering-Theory&Practice, 2021, 41(1):124-133.
[10] Benvenuto D, Giovanetti M, Vassallo L, et al. Application of the ARIMA model on the COVID-2019 epidemic dataset[J]. Data in Brief, 2020, 29:105340.
[11] Singh S, Parmar K S, Kumar J, et al. Development of new hybrid model of discrete wavelet decomposition and autoregressive integrated moving average (ARIMA) models in application to one month forecast the casualties cases of COVID-19[J]. Chaos, Solitons&Fractals, 2020, 135:10986.
[12] Alberti T, Faranda D. On the uncertainty of real-time predictions of epidemic growths:A COVID-19 case study for China and Italy[J]. Communications in Nonlinear Science and Numerical Simulation, 2020, 90:105372.
[13] Liu Y, Gu Z L, Xia S, et al. What are the underlying transmission patterns of COVID-19 outbreak?An age-specific social contact characterization[J]. EClinicalMedicine, 2020, 22:100354.
[14] Wang L, Li J, Guo S, et al. Real-time estimation and prediction of mortality caused by COVID-19 with patient information based algorithm[J]. Science of the Total Environment, 2020, 727:138394.
[15] Doornik J A, Castle J L, Hendry D F. Short-term forecasting of the coronavirus pandemic[J]. International Journal of Forecasting, 2022, 38(2):453-466.
[16] Hasan N. A Methodological approach for predicting COVID-19 epidemic using EEMD-ANN hybrid model[J]. Internet of Things, 2020, 11:100228.
[17] 王志心,刘治,刘兆军.基于机器学习的新型冠状病毒(COVID-19)疫情分析及预测[J].生物医学工程研究, 2020, 39(1):1-5. Wang Z X, Liu Z, Liu Z J. COVID-19 analysis and forecast based on machine learning[J]. Journal of Biomedical Engineering Research, 2020, 39(1):1-5.
[18] Wu J T, Leung K, Lam T T, et al. Nowcasting epidemics of novel pathogens:Lessons from COVID-19[J]. Nature Medicine, 2021, 27(3):388-395.
[19] Al-Rousan N, Al-Najjar H. Nowcasting and forecasting the spreading of novel coronavirus 2019-nCoV and its association with weather variables in 30 Chinese provinces:A case study[J]. Available at SSRN 3537084, 2020.
[20] Wang L, Wang G, Gao L, et al. Spatiotemporal dynamics, nowcasting and forecasting of COVID-19 in the United States[J]. arXiv preprint arXiv:2004.14103, 2020.
[21] Gao Y, Luo M F, Zou G H. Forecasting with model selection or model averaging:A case study for monthly container port throughput[J]. Transportmetrica, 2016, 12(4):366-384.
[22] Shen S, Li G, Song H. Combination forecasts of international tourism demand[J]. Annals of Tourism Research, 2011, 38(1):72-89.
[23] Sun Y Y, Hong Y M, Lee T, et al. Time-varying model averaging[J]. Journal of Econometrics, 2020. https://economics.ucr.edu/repec/ucr/wpaper/202001.pdf.
[24] Sun Y, Zhang J, Li X, et al. Forecasting tourism demand with a new time-varying forecast averaging approach[J]. Journal of Travel Research, 2021. doi:10.1177/00472875211061206.
[25] 张健,孙玉莹,张新雨,等.基于时变模型平均方法的我国航空客运量预测[J].系统工程理论与实践, 2020, 40(6):153-163.Zhang J, Sun Y Y, Zhang X Y, et al. Time-varying forecast averaging for air passengers in China[J]. Systems Engineering-Theory&Practice, 2020, 40(6):153-163.
[26] Wang S Y. TEI@I:A new methodology for studying complex systems[C]//The International Workshop on Complexity Science, Tsukuba, Japan, 2004.
[27] 黄安强,肖进,汪寿阳.一个基于集成情境知识的组合预测方法[J].系统工程理论与实践, 2011, 31(1):55-65.Huang A Q, Xiao J, Wang S Y. A combined forecasting method integrating contextual knowledge[J]. Systems Engineering-Theory&Practice, 2011, 31(1):55-65.
[28] 张嘉为,索丽娜,齐晓楠,等.基于TEI@I方法论的通货膨胀问题分析与预测[J].系统工程理论与实践, 2010, 30(12):2157-2164.Zhang J W, Suo L N, Qi X N, et al. Inflation forecasting method based on TEI@I methodology[J]. Systems Engineering-Theory&Practice, 2010, 30(12):2157-2164.
[29] 张茜,吴超,乔晗,等.基于TEI@I方法论的中国季播电视综艺节目收视率预测[J].系统工程理论与实践, 2016, 36(11):2905-2914.Zhang Q, Wu C, Qiao H, et al. Forecasting audience ratings of China's seasonal entertainment TV shows based on TEI@I methodology[J]. Systems Engineering-Theory&Practice, 2016, 36(11):2905-2914.
[30] Hethcote H W. The mathematics of infectious diseases[J]. SIAM Review, 2000, 42(4):599-653.
[31] Anderson R M, May R M. Infectious diseases of humans:Dynamics and control[M]. Oxford University Press, 1992.
[32] Hethcote H W. A thousand and one epidemic models[M]. Frontiers in Mathematical Biology. Springer, Berlin, Heidelberg, 1994:504-515.
[33] Keeling M J, Rohani P. Modeling infectious diseases in humans and animals[M]. Princeton University Press, 2011.
[34] Liu X, Stechlinski P. Infectious disease models with time-varying parameters and general nonlinear incidence rate[J]. Applied Mathematical Modelling, 2012, 36(5):1974-1994.
[35] Grassly N C, Fraser C. Seasonal infectious disease epidemiology[J]. Proceedings of the Royal Society B:Biological Sciences, 2006, 273(1600):2541-2550.
[36] Dowell S F. Seasonal variation in host susceptibility and cycles of certain infectious diseases[J]. Emerging Infectious Diseases, 2001, 7(3):369.
[37] Wang X, Liu X, Xu W, et al. Stochastic dynamics of HIV models with switching parameters and pulse control[J]. Journal of the Franklin Institute, 2015, 352(7):2765-2782.
[38] Zaman G, Khan A. Dynamical aspects of an age-structured SIR endemic model[J]. Computers&Mathematics with Applications, 2016, 72(6):1690-1702.
[39] He Y, Gao S, Xie D. An SIR epidemic model with time-varying pulse control schemes and saturated infectious force[J]. Applied Mathematical Modelling, 2013, 37(16-17):8131-8140.
[40] Cai Z W. Trending time-varying coefficient time series models with serially correlated errors[J]. Journal of Econometrics, 2007, 136(1):163-188.
[41] 涂云东,汪思韦.函数型核加权估计法及其在经济学中的应用[J].系统工程理论与实践, 2019, 39(4):839-853.Tu Y D, Wang S W. Functional kernel-weighted least square estimation and its applications in economics[J]. Systems Engineering-Theory&Practice, 2019, 39(4):839-853.
[42] Sun H, Qiu Y, Yan H, et al. Tracking and predicting COVID-19 epidemic in China mainland[J]. Journal of Data Science, 2020, 18(3):455-472.
[43] Morotti S, Grandi E. Logistic regression analysis of populations of electrophysiological models to assess proarrythmic risk[J]. MethodsX, 2017, 4:25-34.
[44] Shipe M E, Deppen S A, Farjah F, et al. Developing prediction models for clinical use using logistic regression:An overview[J]. Journal of Thoracic Disease, 2019, 11(4):S574.
[45] Muller C J, MacLehose R F. Estimating predicted probabilities from logistic regression:Different methods correspond to different target populations[J]. International Journal of Epidemiology, 2014, 43(3):962-970.
[46] Ostertagová E. Modelling using polynomial regression[J]. Procedia Engineering, 2012, 48:500-506.
[47] Roy S, Bhunia G S, Shit P K. Spatial prediction of COVID-19 epidemic using ARIMA techniques in India[J]. Modeling Earth Systems and Environment, 2021, 7(2):1385-1391.
[48] Ekum M, Ogunsanya A. Application of hierarchical polynomial regression models to predict transmission of COVID-19 at global level[J]. International Journal of Clinical Biostatistics and Biometrics, 2020, 6(1):27.
[49] Aslam M. Using the kalman filter with Arima for the COVID-19 pandemic dataset of Pakistan[J]. Data in Brief, 2020, 31:105854.
[50] Singh R K, Rani M, Bhagavathula A S, et al. Prediction of the COVID-19 pandemic for the top 15 affected countries:Advanced autoregressive integrated moving average (ARIMA) model[J]. JMIR Public Health and Surveillance, 2020, 6(2):e19115.
[51] Chen B, Hong Y. Testing for smooth structural changes in time series models via nonparametric regression[J]. Econometrica, 2012, 80(3):1157-1183.
[52] Durbin J, Koopman S J. Time series analysis by state space models[M]. Oxford:Oxford University Press, 2001.
[53] Arisoy I, Ozturk I. Estimating industrial and residential electricity demand in Turkey:A time varying parameter approach[J]. Energy, 2014, 66(3):959-964.
[54] Hansen P R, Lunde A, Nason J M. The model confidence set[J]. Econometrica, 2011, 79(2):453-497.
[55] Samuels J D, Sekkel R M. Model confidence sets and forecast combination[J]. International Journal of Forecasting, 2017, 33(1):48-60.
[56] Quaedvlieg R. Multi-horizon forecast comparison[J]. Journal of Business&Economic Statistics, 2021, 39(1):40-53.
[57] Aganin A. Forecast comparison of volatility models on Russian stock market[J]. Applied Econometrics, 2017, 48:63-84.
[58] Zhang Y, Wei Y, Zhang Y, et al. Forecasting oil price volatility:Forecast combination versus shrinkage method[J]. Energy Economics, 2019, 80:423-433.
[59] Yang K, Tian F, Chen L, et al. Realized volatility forecast of agricultural futures using the HAR models with bagging and combination approaches[J]. International Review of Economics&Finance, 2017, 49:276-291.
[60] Li T, Hu Z, Jia Y, et al. Forecasting crude oil prices using ensemble empirical mode decomposition and sparse Bayesian learning[J]. Energies, 2018, 11(7):1882.
[61] 杜本峰.基于Bootstrap方法的风险度量模型及其实证分析-关于机构投资者风险度量方法的探讨[J].统计研究, 2004(1):49-54.Du B F. The risk measuring model based on bootstrap method and the experimental analysis-An approach to the method to measure the risk of institutional investor[J]. Statistical Research, 2004(1):49-54.
[62] 龚旭,曹杰,文凤华,等.基于杠杆效应和结构突变的HAR族模型及其对股市波动率的预测研究[J].系统工程理论与实践, 2020, 40(5):1113-1133.Gong X, Cao J, Wen F H, et al. The HAR-type models with leverage and structural breaks and their applications to the volatility forecasting of stock market[J]. Systems Engineering-Theory&Practice, 2020, 40(5):1113-1133.
[63] 王霞,付中昊,洪永淼,等.基于非参数回归的金融传染检验[J].系统工程理论与实践, 2020, 40(6):1398-1418.Wang X, Fu Z H, Hong Y M, et al. Nonparametric regression based testing for financial contagion[J]. Systems Engineering-Theory&Practice, 2020, 40(6):1398-1418.

Funding

National Natural Science Foundation of China (72073126,72091212);Econometric Modeling and Economic Policy Research Project of Basic Science Center of National Natural Science Foundation of China (71988101);National Energy Group's Top Ten Soft Projects in 2021"Energy System Model Construction and Outlook Research of China's Energy"(GJNY-21-141)
PDF(1188 KB)

Accesses

Citation

Detail
Sections
Recommended

/

〈 〉