Evaluating the forecasting performance of ensemble sub-epidemic frameworks and other time series models for the 2022–2023 mpox epidemic

Royal Society Open Science, Journal Year: 2024, Volume and Issue: 11(7)

Published: July 1, 2024

During the 2022-2023 unprecedented mpox epidemic, near real-time short-term forecasts of epidemic's trajectory were essential in intervention implementation and guiding policy. However, as case levels have significantly decreased, evaluating model performance is vital to advancing field epidemic forecasting. Using laboratory-confirmed data from Centers for Disease Control Prevention Our World Data teams, we generated retrospective sequential weekly Brazil, Canada, France, Germany, Spain, United Kingdom, States at global scale using an auto-regressive integrated moving average (ARIMA) model, generalized additive simple linear regression, Facebook's Prophet well sub-epidemic wave n-sub-epidemic modelling frameworks. We assessed forecast mean squared error, absolute weighted interval scores, 95% prediction coverage, skill scores Winkler scores. Overall, framework outcompeted other models across most locations forecasting horizons, with unweighted ensemble performing best frequently. The spatial-wave frameworks considerably improved relative ARIMA (greater than 10%) all metrics. Findings further support epidemics emerging re-emerging infectious diseases.

Language: Английский

Short-Term Predictions of the Trajectory of Mpox in East Asian Countries, 2022–2023: A Comparative Study of Forecasting Approaches

Mathematics, Journal Year: 2024, Volume and Issue: 12(23), P. 3669 - 3669

Published: Nov. 23, 2024

The 2022–2023 mpox outbreak exhibited an uneven global distribution. While countries such as the UK, Brazil, and USA were most heavily affected in 2022, many Asian countries, specifically China, Japan, South Korea, Thailand, experienced later, 2023, with significantly fewer reported cases relative to their populations. This variation timing scale distinguishes outbreaks these from those first wave. study evaluates predictability of smaller case counts using popular epidemic forecasting methods, including ARIMA, Prophet, GLM, GAM, n-Sub-epidemic, Sub-epidemic Wave frameworks. Despite fact that ARIMA GAM models performed well for certain prediction windows, results generally inconsistent highly dependent on country, i.e., dataset, interval length. In contrast, n-Sub-epidemic Ensembles demonstrated more reliable robust performance across different datasets predictions, indicating effectiveness this model small its utility early stages future pandemics.

Language: Английский