tetano
Editor, Senior Moderator
J Adv Res
. 2026 Apr 4:S2090-1232(26)00288-2.
doi: 10.1016/j.jare.2026.04.009. Online ahead of print.
Dynamic feature selection improves influenza forecasting accuracy and generalization across countries
Jingyi Liang[SUP] 1 [/SUP], Zhiqi Zeng[SUP] 2 [/SUP], Kai Liao[SUP] 3 [/SUP], Min Liao[SUP] 4 [/SUP], Zhonghao Fang[SUP] 1 [/SUP], Lesi Kong[SUP] 5 [/SUP], Jianlin Feng[SUP] 1 [/SUP], Na Yu[SUP] 4 [/SUP], Zhengshi Lin[SUP] 6 [/SUP], Chitin Hon[SUP] 7 [/SUP], Arlindo Oliveira[SUP] 8 [/SUP], Zifeng Yang[SUP] 9 [/SUP]
Affiliations
Introduction: The differentiation in epidemic patterns and multiple influencing factors pose significant challenges to influenza forecasting, highlighting the need for novel methods to improve predictive accuracy and cross-regional generalizability.
Objectives: This study aims to develop an adaptive feature selection model named AdaFluDR to address the time-varying nature of influenza transmission drivers across different periods and regions.
Methods: AdaFluDR integrates the SpaceTime and Crossformer models and utilizes a correlation-driven mechanism. This mechanism constructs a comprehensive score by integrating the temporal, frequency, and time domain information of features, and dynamically adjusts feature processing pathways based on this score. Subsequently, a multilayer perceptron (MLP) will be employed to model the nonlinear mapping relationship between the integrated features and the target variable for prediction generation.
Results: The AdaFluDR model outperforms traditional methods and other machine learning approaches, demonstrating robust predictive performance across multiple forecasting horizons (1-4 weeks), and strong generalization ability across the United States, Canada, and Portugal.
Conclusion: Our study provides a novel and practical framework for forecasting influenza activity with reliable accuracy and cross-national applicability, providing a valuable tool for improving global epidemic preparedness and response strategies.
Keywords: Adaptive decisionmodel; Dynamic featureextraction; Full-period modeling; Influenza forecasting; Multi-region and multi-periodforecasting.
. 2026 Apr 4:S2090-1232(26)00288-2.
doi: 10.1016/j.jare.2026.04.009. Online ahead of print.
Dynamic feature selection improves influenza forecasting accuracy and generalization across countries
Jingyi Liang[SUP] 1 [/SUP], Zhiqi Zeng[SUP] 2 [/SUP], Kai Liao[SUP] 3 [/SUP], Min Liao[SUP] 4 [/SUP], Zhonghao Fang[SUP] 1 [/SUP], Lesi Kong[SUP] 5 [/SUP], Jianlin Feng[SUP] 1 [/SUP], Na Yu[SUP] 4 [/SUP], Zhengshi Lin[SUP] 6 [/SUP], Chitin Hon[SUP] 7 [/SUP], Arlindo Oliveira[SUP] 8 [/SUP], Zifeng Yang[SUP] 9 [/SUP]
Affiliations
- PMID: 41942047
- DOI: 10.1016/j.jare.2026.04.009
Introduction: The differentiation in epidemic patterns and multiple influencing factors pose significant challenges to influenza forecasting, highlighting the need for novel methods to improve predictive accuracy and cross-regional generalizability.
Objectives: This study aims to develop an adaptive feature selection model named AdaFluDR to address the time-varying nature of influenza transmission drivers across different periods and regions.
Methods: AdaFluDR integrates the SpaceTime and Crossformer models and utilizes a correlation-driven mechanism. This mechanism constructs a comprehensive score by integrating the temporal, frequency, and time domain information of features, and dynamically adjusts feature processing pathways based on this score. Subsequently, a multilayer perceptron (MLP) will be employed to model the nonlinear mapping relationship between the integrated features and the target variable for prediction generation.
Results: The AdaFluDR model outperforms traditional methods and other machine learning approaches, demonstrating robust predictive performance across multiple forecasting horizons (1-4 weeks), and strong generalization ability across the United States, Canada, and Portugal.
Conclusion: Our study provides a novel and practical framework for forecasting influenza activity with reliable accuracy and cross-national applicability, providing a valuable tool for improving global epidemic preparedness and response strategies.
Keywords: Adaptive decisionmodel; Dynamic featureextraction; Full-period modeling; Influenza forecasting; Multi-region and multi-periodforecasting.