Zhu Q, Zhu M, Cai Y, Xiang J, Li S, Zhu B, Zhu M,. A stacked ensemble model with NNLS-based weighting for influenza forecasting: a case study of Anhui Province, China. Front Public Health. 2026 May 5;14:1806095
Background: Influenza poses a significant global public health threat, with its pandemic potential and seasonal variability presenting formidable challenges to prediction accuracy. This study leverages high-quality weekly data (incidence rates, viral subtypes, and meteorological indicators) from the provincial influenza surveillance system in Anhui Province, eastern China, spanning 2015-2025. A multi-source data fusion model was developed to overcome the limitations of traditional methods in modeling nonlinear transmission dynamics and multi-factor synergistic effects.
Methods: Single models were constructed using ARIMA, Prophet, and XGBoost, then stacked into an interpretable ensemble model (Stacked-NNLS) using non-negative least squares (NNLS). Performance was comprehensively evaluated using R 2 (explained variance), RMSE (root mean square error), MAE (mean absolute error), and MAPE (mean absolute percentage error).
Results: ARIMA exhibits poor fit for non-stationary sequences (training set R 2 = -3.66; test set R 2 = 0.03). Prophet effectively captures long-term trends (training/test set R 2 = 0.38/0.88). XGBoost shows overfitting (training/test set R 2 = 0.99/0.74). The Stacked-NNLS model demonstrated significantly superior robustness (training/test R 2 = 0.94/0.94), outperforming baseline models across all metrics.
Conclusion: By integrating statistical, seasonal, and nonlinear modeling approaches, Stacked-NNLS demonstrated robust predictive performance in capturing influenza trends, seasonal fluctuations, and complex interactions among multiple factors, suggesting its potential utility for infectious disease early warning and public health decision-making.
Methods: Single models were constructed using ARIMA, Prophet, and XGBoost, then stacked into an interpretable ensemble model (Stacked-NNLS) using non-negative least squares (NNLS). Performance was comprehensively evaluated using R 2 (explained variance), RMSE (root mean square error), MAE (mean absolute error), and MAPE (mean absolute percentage error).
Results: ARIMA exhibits poor fit for non-stationary sequences (training set R 2 = -3.66; test set R 2 = 0.03). Prophet effectively captures long-term trends (training/test set R 2 = 0.38/0.88). XGBoost shows overfitting (training/test set R 2 = 0.99/0.74). The Stacked-NNLS model demonstrated significantly superior robustness (training/test R 2 = 0.94/0.94), outperforming baseline models across all metrics.
Conclusion: By integrating statistical, seasonal, and nonlinear modeling approaches, Stacked-NNLS demonstrated robust predictive performance in capturing influenza trends, seasonal fluctuations, and complex interactions among multiple factors, suggesting its potential utility for infectious disease early warning and public health decision-making.
See Also:
Latest articles in those days:
- The evolution and transmission dynamics of H3N2 influenza virus identified in a respiratory disease outbreak in companion dogs in Alabama, USA 6 hours ago
- Structural and functional characterization of hemagglutinin E190D and Q226H minor variants in human H5N1 A/British Columbia/PHL-2032/2024 6 hours ago
- Spatial risk factors associated with outbreaks of highly pathogenic avian influenza in Pennsylvania’s commercial poultry farms 6 hours ago
- [preprint]H9N2 avian influenza virus in Bangladesh: three genetically distinct co-circulating subgroups retain a largely conserved antigenic phenotype 6 hours ago
- [preprint]Endogenous serine proteases in MDCK sublines support trypsin-independent influenza A virus replication 6 hours ago
[Go Top] [Close Window]


