EVALUATION OF DIFFERENT IMPUTATION METHODS FOR HANDLING MISSING AIR POLLUTION DATA IN BURSA, TURKEY

Abstract

Air pollution data integrity is paramount for effective environmental management and public health protection. This study aims to address the issue of data integrity related to air pollution in Bursa and evaluates various advanced imputation techniques to effectively complete missing data. Five different methods, including linear regression, K-Nearest Neighbors (KNN), decision trees, random forests, and linear interpolation, are compared for filling gaps in the PM_2.5 dataset. Our findings demonstrate that these advanced imputation techniques enhance the completeness and accuracy of air pollution data.

The significance of our study underscores the critical role of accurately and reliably completing missing air pollution data in environmental management and public health. These findings can guide researchers in selecting the most appropriate imputation method for the analysis of air pollution data. Additionally, this study establishes a foundation for addressing the issue of missing data in future research and the formulation of environmental policies.

This study underscores the vital role of imputation methods in ensuring data reliability, contributing to more effective environmental monitoring, and safeguarding public health.