Title: Wavelets and hybrid optimised SVM with random forest-based pollution forecasting

Authors: Zaheer Abbas; Princess Raina

Addresses: Department of Mathematical Sciences, Baba Ghulam Shah Badshah University, Rajouri, India ' Department of Mathematical Sciences, Baba Ghulam Shah Badshah University, Rajouri, India

Abstract: Due to its detrimental effects on human health, information about meteorological pollutants including CO, NO₂, SO₂, and dust is becoming more and more crucial. In all nations' urban areas, this is particularly true. The instantaneous registration provided by the automatic measurements of these pollutants' concentrations serves as the foundation for the calculation of averaged values. The key issue is early pollution forecasting in order to warn or notify the local population of the impending risk. In this work, machine learning and wavelet decomposition are used to forecast daily air pollution. This research provides the forecasting strategy, utilising the hybrid random forrest with optimised support vector machines (HRFOpSV), depend on the collected data of NO₂, CO, SO₂, and dust, for the past years, and real weather conditions like humidity, wind, temperature and pressure.

Keywords: pollution; machine learning; forecasting; decomposition; human health.

DOI: 10.1504/IJGW.2024.135368

International Journal of Global Warming, 2024 Vol.32 No.1, pp.81 - 101

Received: 13 Apr 2023
Accepted: 07 Jun 2023
Published online: 06 Dec 2023 *