Symptom-Based COVID19 Screening Model Combined with Surveillance Information

Area of Science:

• Epidemiology
• Machine Learning
• Public Health

Background:

• The rapid increase in COVID-19 cases necessitates advanced screening methods.
• Effective COVID-19 screening is crucial for public health management and control.

Purpose of the Study:

• To evaluate the impact of surveillance indices on the performance of machine learning models for COVID-19 screening.
• To compare the accuracy of models trained with and without COVID-19 surveillance data.

Main Methods:

• Training and evaluating machine learning models using the Israel COVID-19 dataset.
• Comparing model performance metrics (AUC scores) with and without the inclusion of surveillance information.

Main Results:

• Models incorporating surveillance indices achieved higher AUC scores (0.8478±0.0037) compared to those without (0.8062±0.005).
• A statistically significant improvement in model performance was observed when surveillance data was utilized.

Conclusions:

• Incorporating surveillance data significantly enhances the accuracy of machine learning-based COVID-19 screening.
• Surveillance indices are valuable components for developing more effective COVID-19 detection and monitoring tools.