RESUMO
Malaria is the most common cause of death among the parasitic diseases. Malaria continues to pose a growing threat to the public health and economic growth of nations in the tropical and subtropical parts of the world. This study aims to address this challenge by developing a predictive model for malaria outbreaks in each district of The Gambia, leveraging historical meteorological data. To achieve this objective, we employ and compare the performance of eight machine learning algorithms, including C5.0 decision trees, artificial neural networks, k-nearest neighbors, support vector machines with linear and radial kernels, logistic regression, extreme gradient boosting, and random forests. The models are evaluated using 10-fold cross-validation during the training phase, repeated five times to ensure robust validation. Our findings reveal that extreme gradient boosting and decision trees exhibit the highest prediction accuracy on the testing set, achieving 93.3% accuracy, followed closely by random forests with 91.5% accuracy. In contrast, the support vector machine with a linear kernel performs less favorably, showing a prediction accuracy of 84.8% and underperforming in specificity analysis. Notably, the integration of both climatic and non-climatic features proves to be a crucial factor in accurately predicting malaria outbreaks in The Gambia.
Assuntos
Surtos de Doenças , Aprendizado de Máquina , Malária , Máquina de Vetores de Suporte , Gâmbia/epidemiologia , Humanos , Malária/epidemiologia , Redes Neurais de Computação , AlgoritmosRESUMO
Although the relationship between the environmental factors, such as weather conditions and air pollution, and COVID-19 case fatality rate (CFR) has been found, the impacts of these factors to which infected cases are exposed at different infectious stages (e.g., virus exposure time, incubation period, and at or after symptom onset) are still unknown. Understanding this link can help reduce mortality rates. During the first wave of COVID-19 in the United Kingdom (UK), the CFR varied widely between and among the four countries of the UK, allowing such differential impacts to be assessed. We developed a generalized linear mixed-effect model combined with distributed lag nonlinear models to estimate the odds ratio of the weather factors (i.e., temperature, sunlight, relative humidity, and rainfall) and air pollution (i.e., ozone, [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text] and [Formula: see text]) using data between March 26, 2020 and September 15, 2020 in the UK. After retrospectively time adjusted CFR was estimated using back-projection technique, the stepwise model selection method was used to choose the best model based on Akaike information criteria and the closeness between the predicted and observed values of CFR. The risk of death reached its maximum level when the low temperature (6 °C) occurred 1 day before (OR 1.59; 95% CI 1.52-1.63), prolonged sunlight duration (11-14 h) 3 days after (OR 1.24; 95% CI 1.18-1.30) and increased [Formula: see text] (19 µg/m3) 1 day after the onset of symptom (OR 1.12; 95% CI 1.09-1.16). After reopening, many COVID-19 cases will be identified after their symptoms appear. The findings highlight the importance of designing different preventive measures against severe illness or death considering the time before and after symptom onset.