A1049
Title: An application of random forests to estimate the reporting delay in COVID-19 cases
Authors: Xuanan Lin - Keio University (Japan) [presenting]
Hiroshi Shiraishi - Keio University (Japan)
Abstract: Accurate forecasting of COVID-19 cases is paramount for effective public health planning and response. However, the presence of reporting delays complicates this task, leading to underestimation or misrepresentation of the true disease burden. A novel approach leveraging random forests is introduced to evaluate the effect of reporting delays on COVID-19 case counts. By integrating historical case data with features indicative of reporting lags, such as infection time, patient density, and development periods, the random forests model is adapted to capture complex relationships and nonlinear effects inherent in the reporting process. Model performance is rigorously evaluated using the mean squared error (MSE), providing a quantitative measure of predictive accuracy. The application of random forests unveils insights into the temporal dynamics of reporting delays and their implications for epidemic surveillance and control efforts. The utility of machine learning methodologies, particularly the random forests, is underscored in unravelling the intricacies of infectious disease surveillance and informing evidence-based public health policies to mitigate the impact of reporting delays on COVID-19 case estimation.
