Predicting dengue importation into Europe, using machine learning and model-agnostic methods

Abstract

The geographical spread of dengue is a global public health concern. This is largely mediated by the importation of dengue from endemic to non-endemic areas via the increasing connectivity of the global air transport network. The dynamic nature and intrinsic heterogeneity of the air transport network make it challenging to predict dengue importation. Here, we explore the capabilities of state-of-the-art machine learning algorithms to predict dengue importation. We trained four machine learning classifiers algorithms, using a 6-year historical dengue importation data for 21 countries in Europe and connectivity indices mediating importation and air transport network centrality measures. Predictive performance for the classifiers was evaluated using the area under the receiving operating characteristic curve, sensitivity, and specificity measures. Finally, we applied practical model-agnostic methods, to provide an in-depth explanation of our optimal model’s predictions on a global and local scale. Our best performing model achieved high predictive accuracy, with an area under the receiver operating characteristic score of 0.94 and a maximized sensitivity score of 0.88. The predictor variables identified as most important were the source country’s dengue incidence rate, population size, and volume of air passengers. Network centrality measures, describing the positioning of European countries within the air travel network, were also influential to the predictions. We demonstrated the high predictive performance of a machine learning model in predicting dengue importation and the utility of the model-agnostic methods to offer a comprehensive understanding of the reasons behind the predictions. Similar approaches can be utilized in the development of an operational early warning surveillance system for dengue importation.

Impacts of El Niño Southern Oscillation on the dengue transmission dynamics in the Metropolitan Region of Recife, Brazil

“…Therefore, determining epidemic synchronism and regularity at the metropolitan scale shows how the disease evolves and can be connected between municipalities. However, it is imperative to expand the analysis to other scales because viruses and vectors can be transported from all distances [61][62][63] .…”

An Inverse Relation between Hyperglycemia and Skeletal Muscle Mass Predicted by Using a Machine Learning Approach in Middle-Aged and Older Adults in Large Cohorts

“…Unlike traditional computer algorithms that produce results when commands are executed, machine learning algorithms can construct and train algorithms to process data and obtain results in required forms [ 3 ]. In the medical field, machine learning-based predictive algorithms have been developed to analyze metabolic factors related to various diseases [ 4 , 5 , 6 ], and in the present study, we used machine learning algorithms to create regression models that predict SMM and FM. In men and women, muscle mass accounts for 47–60% of body weight and maintains energy expenditure throughout the body [ 7 ].…”

“…In older adults, SMM reduction is linked to increases in fat mass [ 5 ]. Furthermore, the prevalence of obesity is increasing rapidly [ 6 ], and obesity increases the risk of various metabolic diseases, such as diabetes and cardiovascular disease [ 6 ]. In addition, obesity and SMM loss can act synergistically to cause more severe health problems [ 7 , 8 ].…”

Prediction of Aedes aegypti breeding distribution through spatiotemporal analysis and machine learning: A case study in Recife, Pernambuco

“…The relationship between adequacy of temperature and precipitation was considered responsible for most of the occurrence of the disease, thus offering a statistical explanation for the spatiotemporal variability in its transmission. Salami et al (2020) predicted the importation of Dengue cases in Europe in 21 countries, through machine learning and independent model methods. Four classification algorithms were trained: partial least squares (pls), generalized linearized models of loop and elastic network (glmnet), random forest (randomForest), extreme gradient impulse (xgboost), using historical data from 06 years (2010)(2011)(2012)(2013)(2014)(2015).…”