Intelligent air pollution prediction algorithm-based optimized random forest regression for reducing asthmatic attacks

Introduction: Air pollution can trigger the attack in asthmatic patients if uncontrolled. Previous works focused on controlling pollution by proposing algorithms to predict air pollution. While these prediction algorithms save patients from attack triggers, they have limitations such as prediction accuracy, mathematical complexity, and lack of adequate patient notification systems. Materials and methods: This study proposed a novel Intelligent Air Pollution Prediction (IAPP) algorithm based on optimizing Random Forest Regression (RFR) to predict air pollution and send an alert message to the patient and hospital in real time. Meanwhile, IAPP utilized reliable data from Internet of Things (IoT)-based air pollution detection nodes. The performance of IAPP was evaluated in a real-world environment during the peak pollutant season to test the prediction accuracy of air pollution. Results: Results showed that the proposed IAPP achieved a high prediction accuracy of 99.98% with an R-squared value of 0.99. This demonstrated that the IAPP algorithm based on the RFR model can effectively protect asthmatic patients from attack triggers. Conclusion: As a result, the IAPP algorithm reduces hospital visits during high pollution and enables patients to complete their daily activities without obstacles or absence.