收藏切换
Noise reduction and prediction performance optimization of influenza surveillance data based on mechanism modeling and data assimilation
收藏切换
Xin LI1, Yu-xing TIAN1, Heng YUAN2, Tao ZHANG1
Modern Preventive Medicine | 2025, 52(22) : 4046 - 4051
Less
收藏切换
Modern Preventive Medicine | 2025, 52(22): 4046-4051
• Epidemiology and Statistical Methods Advances •
Noise reduction and prediction performance optimization of influenza surveillance data based on mechanism modeling and data assimilation
Full
Xin LI1, Yu-xing TIAN1, Heng YUAN2, Tao ZHANG1
Affiliations
  • Department of Epidemiology and Health Statistics, West China School of Public Health and West China Fourth Hospital, Sichuan University, Chengdu, Sichuan 610041, China
Published: 2025-11-25 doi: 10.20043/j.cnki.MPM.202506072
Outline
收藏切换
Objective

To improve the structural quality of influenza surveillance data and enhance the performance of predictive models by leveraging an epidemic dynamics-informed data assimilation approach.

Methods

This study constructed a data assimilation framework that combined a Susceptible–Infectious–Recovered–Susceptible(SIRS)model with the Ensemble Adjustment Kalman Filter(EAKF)to assimilate weekly Goldstein indices(Influenza-like Illness Percentage × Positive Rate)from 21 cities in Sichuan Province. The signal-to-noise ratio(SNR)was used to evaluate the effect of data structure optimization, and predictive performance before and after assimilation was assessed using three forecasting models: ARIMA, RNN, and LSTM. Linear mixed-effect models were applied to quantify the impact of assimilation on SNR and root mean squared error(RMSE). Statistical significance was defined as P<0.05.

Results

After applying the SIRS-EAKF assimilation framework, the SNR significantly increased in 18 cities(85.7%), with an average improvement of 3.26(P<0.01). Across all three prediction models, RMSE were significantly reduced in the assimilated data group(all P<0.05).

Conclusion

The SIRS-EAKF framework effectively suppressed the noise and enhanced the expression of the dominant epidemic signal through incorporating dynamical mechanisms. Without altering the structure of the prediction model, it significantly improved both the quality of influenza surveillance data and the predictive accuracy. This approach provides a robust methodological foundation for real-time epidemic forecasting and public health decision-making.

Influenza surveillance  /  Data assimilation  /  SIRS model  /  EAKF  /  Time series forecasting
李鑫, 田宇星, 袁珩, 张韬. 基于机制建模与数据同化的流感监测数据降噪与预测性能优化研究. 现代预防医学, 2025 , 52 (22) : 4046 -4051 . DOI: 10.20043/j.cnki.MPM.202506072
Xin LI, Yu-xing TIAN, Heng YUAN, Tao ZHANG. Noise reduction and prediction performance optimization of influenza surveillance data based on mechanism modeling and data assimilation[J]. Modern Preventive Medicine, 2025 , 52 (22) : 4046 -4051 . DOI: 10.20043/j.cnki.MPM.202506072
Year 2025 volume 52 Issue 22
47
21
Cite this Article
BibTeX
Article Info
doi: 10.20043/j.cnki.MPM.202506072
  • Receive Date:2025-06-05
  • Online Date:2026-10-29
  • Published:2025-11-25
Article Data
Affiliations
History
  • Received:2025-06-05
Funding
Affiliations
    Department of Epidemiology and Health Statistics, West China School of Public Health and West China Fourth Hospital, Sichuan University, Chengdu, Sichuan 610041, China
References
Share
https://castjournals.cast.org.cn/joweb/xdyfyx/EN/10.20043/j.cnki.MPM.202506072
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)
属
Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT