Optimized MSVM-RFE Model for Enhanced Urban Air Quality Management and Decision Support

A useful instrument for keeping urban air quality at an acceptable level is the Urban Air Quality Management Plan (UAQMP). However, UAQM techniques are tailored to meet the unique requirements of each country. In developed countries, regulatory management systems often include UAQMPs. However, developing nations still don't know how to implement UAQMPs that are both effective and practicable to limit the worsening of their urban air quality. The approach consists of three phases, which include preprocessing, feature extraction, and training the model. As a technique for offline frequency-domain filtering, the Hampel filter is employed in preprocessing to eliminate spectral outliers. There are three types of characteristics used in feature extraction: those pertaining to traffic, human mobility, and road networks. It employed an Improved MSVM-RFE for the model's training. The proposed approach surpasses RFE and MSVM with an average accuracy of 93.57%.