Abstract
WHAT IS ALREADY KNOWN ABOUT THIS TOPIC? Current evidence regarding which clinical manifestations best predict influenza requires refinement, particularly considering regional variations in disease presentation and their importance for early diagnosis and surveillance. WHAT IS ADDED BY THIS REPORT? The optimal machine learning model identified key influenza predictors, including epidemiological characteristics, critical symptoms and signs, and age. Based on this model, we introduced a new influenza-like illness (ILI) definition characterized by fever (≥37.9 °C) with either cough or rhinorrhea. WHAT ARE THE IMPLICATIONS FOR PUBLIC HEALTH PRACTICE? These findings provide evidence-based clinical manifestations for influenza prediction and offer an optimized definition of ILI for improved surveillance and early detection.