A Hybrid Forecasting Model Combining Autoregressive Integrated Moving Average and Gaussian Process Regression

Jiale He

doi:10.6981/FEM.202506_6(6).0013

Authors

Jiale He

DOI:

https://doi.org/10.6981/FEM.202506_6(6).0013

Keywords:

ARIMA; Gaussian Process Regression; Hybrid Forecasting Model; Time Series Analysis; Machine Learning.

Abstract

The prediction of population mortality rates is fundamental to economic development, healthcare resource allocation, and the construction of pension systems. Based on the population mortality rate data of China from 1978 to 2022, this paper establishes an ARIMA-GPR hybrid model to forecast the population mortality rate. Initially, ARIMA models with different combinations of p and q are employed to capture the linear trends of the original time series. The five best-performing combinations are selected as input features for the Gaussian Process Regression (GPR) model, which then outputs the predicted values after further fitting. The results indicate that the ARIMA-GPR hybrid model significantly improves the forecasting accuracy compared to the standalone ARIMA model.

Downloads

Download data is not yet available.

References

[1] Yan Zeng, Xi Chen, Yinglu Deng. Innovative Dynamic Population Mortality Prediction and Its Application [J]. Systems Engineering - Theory & Practice, 2016, 36(7): 1710-1718.

[2] Xiaofeng Wu, Yingmei Yang, Yaotong Chen. Combined ARIMA Forecasting Model Based on BP Neural Network Error Correction [J]. Statistics and Decision, 2019, 35(15): 65-68.

[3] Zhikun He, Guangbin Liu, Xijing Zhao, et al. A Survey of Gaussian Process Regression Methods [J]. Control and Decision, 2013, 28(8): 1121-1129+1137.

[4] Zhang G P. Time Series Forecasting Using a Hybrid ARIMA and Neu⁃ ral Network Model [J].Neurocomputing,2003,(50).

[5] Bicong Ci, Pinyi Zhang. Financial time series forecasting based on ARIMA-LSTM model [J]. Statistics & Decision, 2022, 38(11): 145-149.

[6] Jinhai Yao, Jiajun Zou. Construction and numerical simulation of SVM-ARIMA model for CPI prediction [J]. Statistics & Decision, 2022, 38(21): 48-52.

[7] Fengming Zhang, Qian Su, Zhixing Deng, et al. Deformation prediction and application research of metro station foundation pit based on GF-GPR [J]. Journal of Hefei University of Technology (Natural Science Edition), 2025, 48(4): 563-569.

[8] Jinhai Yao. Stock index forecasting research based on ARIMA and information granulation SVR combination [J]. Operations Research and Management Science, 2022, 31(5): 214-220.

[9] Wenting Zong, Zhinong Wei, Guoqiang Sun, et al. Short-term load interval forecasting based on improved Gaussian process regression model [J]. Proceedings of the CSU-EPSA, 2017, 29(8): 22-28.

[10] YU J. State of health prediction of lithium-ion batteries:multiscale logic regression and Gaussian process regression ensemble[J]. Reliability engineering & system safety,2018,174:82-95.