What question did this study set out to answer?

The study aims to develop and evaluate FaXNet, a novel deep learning framework for accurate influenza forecasting.

February 5, 2026Frontiers in Public HealthOpen Access

FaXNet: a frequency-adaptive, explainable, and uncertainty-aware network for influenza forecasting

View Full Paper

Ask AI

Bookmark

Authors

FaXNet demonstrates superior influenza forecasts in China, indicating effective multi-scale modeling techniques.

The study aims to develop and evaluate FaXNet, a novel deep learning framework for accurate influenza forecasting.
Developed FaXNet integrating spectral representation and probabilistic forecasting.
Compiled weekly influenza positivity rates and aligned them with weather data from 2011 to 2023.
Evaluated against various forecasting baselines for 1–4-week-ahead predictions using accuracy and calibration metrics.
FaXNet achieved 1-week-ahead R 2 of 0.9319 in the north and 0.8665 in the south.
4-week-ahead R 2 was 0.4493 (north) and 0.4960 (south).
It showed significant performance improvements over all benchmarks, especially in mitigation of error accumulation.

He et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

February 5, 2026Frontiers in Public HealthOpen Access

View Full Paper

Ask AI

Bookmark

Authors

FaXNet demonstrates superior influenza forecasts in China, indicating effective multi-scale modeling techniques.

The study aims to develop and evaluate FaXNet, a novel deep learning framework for accurate influenza forecasting.
Developed FaXNet integrating spectral representation and probabilistic forecasting.
Compiled weekly influenza positivity rates and aligned them with weather data from 2011 to 2023.
Evaluated against various forecasting baselines for 1–4-week-ahead predictions using accuracy and calibration metrics.
FaXNet achieved 1-week-ahead R 2 of 0.9319 in the north and 0.8665 in the south.
4-week-ahead R 2 was 0.4493 (north) and 0.4960 (south).
It showed significant performance improvements over all benchmarks, especially in mitigation of error accumulation.

He et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark