Spatio-temporal modelling of asthma-prone areas using a machine learning optimized with metaheuristic algorithms

The main aim of this study was to use two metaheuristic optimization algorithms-a genetic algorithm (GA) and a teaching-learning-based optimization (TLBO) algorithm-to determine the optimal parameters of a support vector regression (SVR) model for Spatio-temporal modelling of asthma-prone areas in Tehran, Iran. First, a spatial-temporal database consisting of dependent (872 patients with asthma) and independent data (air pollution, meteorology, distance to park, and street parameters) was created. In the next step, Spatio-temporal modelling and mapping of asthma-prone areas were performed using three models: SVR, SVR-GA, and SVR-TLBO. The highest accuracy of the area under the curve (AUC) of the receiver operating characteristic (ROC) was for SVR-GA (0.806, 0.801, 0.823, and 0.811), SVR-TLBO (0.8, 0.797, 0.81, and 0.803), and SVR (0.786, 0.78, 0.796, and 0.791) models in spring, summer, autumn, and winter, respectively. Autumn, winter, spring, and summer were most accurate in modelling asthma occurrence, respectively.

Automated summary

The aim of this study was to optimize the parameters of a support vector regression (SVR) model for spatio-temporal modeling of asthma-prone areas in Tehran, Iran using genetic algorithm (GA) and teaching-learning-based optimization (TLBO) algorithm. The results showed that the SVR-GA and SVR-TLBO models had higher accuracy than the SVR model in predicting asthma occurrence in different seasons.