Data Sheet 1_A stacked ensemble model with NNLS-based weighting for influenza forecasting:

Weekly influenza surveillance data from Anhui Province, China, spanning 2015 to 2025. The dataset includes incidence rates, viral subtypes, and meteorological indicators. It was used to develop a stacked ensemble forecasting model by Qingqing Zhu.

Use Cases

• Benchmark time-series forecasting models based on weekly incidence and meteorological data.
• Develop ensemble models for infectious disease prediction based on multi-source data fusion.
• Analyze the relationship between influenza trends and environmental factors based on meteorological indicators.
• Evaluate model performance for non-stationary sequences based on metrics like R-squared and RMSE.

Strengths

• Data spans a 10-year period from 2015 to 2025.
• Description includes specific model performance metrics (e.g., Stacked-NNLS test R-squared of 0.94).
• Integrates multiple data sources, including incidence rates, viral subtypes, and meteorological indicators.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data is specific to a single province (Anhui, China), which may limit generalizability.

Provenance

Source

Provincial influenza surveillance system in Anhui Province, China.

Collection Method

Collected as part of a public health surveillance system.

Time Range

2015-2025

Freshness

Last updated 2026-05-05 05:28:37

Geography

Anhui Province, eastern China.