An ensemble model based on early predictors to forecast COVID-19 health care demand in France

Abstract

Significance The COVID-19 pandemic is inducing significant stress on health care structures, which can be quickly saturated with negative consequences for patients. As hospitalization comes late in the infection history of a patient, early predictors—such as the number of cases, mobility, climate, and vaccine coverage—could improve forecasts of health care demand. Predictive models taken individually have their pros and cons, and it is advantageous to combine the predictions in an ensemble model. Here, we design an ensemble that combines several models to anticipate French COVID-19 health care needs up to 14 days ahead. We retrospectively test this model, identify the best predictors of the growth rate of hospital admissions, and propose a promising approach to facilitate the planning of hospital activity.

Holistic One Health Surveillance Framework: Synergizing Environmental, Animal, and Human Determinants for Enhanced Infectious Disease Management

“…161 During the COVID-19 pandemic, epidemic modeling has played a crucial role in assessing risks, predicting its impact on healthcare systems, providing rationale for implementing measures, optimizing vaccine distribution, and influencing various aspects of our response to the virus. 162,163 ACS Infectious Diseases Modern drug discovery has entered the Big Data era due to the vast data sets available for drug candidates. AI, particularly through approaches like deep learning, has become central to innovative modeling, leveraging the dynamic and large nature of these data sets.…”

“…These models incorporate nonlinear equations that describe disease incidence as a function of both susceptible and infected populations, shaping the trajectory of the epidemic . During the COVID-19 pandemic, epidemic modeling has played a crucial role in assessing risks, predicting its impact on healthcare systems, providing rationale for implementing measures, optimizing vaccine distribution, and influencing various aspects of our response to the virus. , …”

An adaptive weight ensemble approach to forecast influenza activity in the context of irregular seasonality

“…We test the following approaches to determine the weights: 1) Simple averaging ensemble (SAE) model: In this ensemble model, we select the top three models based on RMSE at each week t, and then take the unweighted mean of the individual forecast of these 3 models [9] (ie ‫ݓ‬ ௧ =0 if model I is not selected and 1 if it is). 2) Normal Blending Ensemble (NBE) model: In this approach, we fit a LASSO regression of the observed ILI+ up to week t-1 on the predicted ILI+ from different models to obtain the coefficients to generate ensemble forecast for week t [46].…”

“…Forecasting infectious disease activity could inform public health response to outbreaks, such as preparing the increase of hospitalization [2,3]. In regions with stable seasonality such as the continental U.S, forecasting of influenza and COVID-19 could be reliable, with establishment of forecasting hub [4] to generate ensemble forecast based on the forecast from several groups based on several mechanistic, statistical, machine learning and deep learning approaches [4][5][6][7][8][9][10].…”

“…Using Hong Kong as example, based on >20 years of data (1998-2019), we develop and evaluate several statistical, machine learning and deep learning approach. We further explore several ensemble forecast approaches that previously show superior performance in other diseases [4, 9, 17, 19] or geographics [20, 21]. We propose the idea of allowing model weights to be updated dynamically over time to better capture the rapid change of influenza activity in Hong Kong.…”