A Deep Learning Approach for Prediction of Respiratory Disease from Air Quality

Air pollution is a dirty air due to the presence of substances harmful or toxic to human health. It is associated with the respiratory diseases that is a very serious problem and is increasing in severity every year. Therefore, it is necessary to study and analyze data from patients with respiratory diseases. This is because the hospitals must to be prepared to accommodate the increasing number of patients in the future.The objective of this study was to analyze the relationship between air quality and the number of respiratory patients. The air quality data was collected at different times. To determine the relationship of air quality, disease and the number of patients receiving respiratory treatment during the same period. In addition, to create a computational forecasting model to be able to predict respiratory disease from air quality.The study focused on deep learning (DL) method approach as it has the advantage of large input data. The DL model were compared to others method consisting of k-nearest neighbors, linear regression, decision tree and neural network. Performance assessments were performed using a five-fold cross validation method which used root mean square error (RMSE) for comparison. The results shows that the DL model provides the best performance.