Analyzing Urban Air Quality through Machine Learning and Geographic Context

This study employs machine learning to analyze the relationship between air pollutants and climatic and geographical factors in Bandar Lampung, Indonesia, using data from January 2022 to November 2023. A strong correlation was found between PM2.5, PM10 (r=0.99), and seasonal increases in O3, SO2, and NO2 during the dry season, highlighting the influence of weather and geography on air quality. Utilizing the CRISP-DM framework, the study offers insights for urban planning and pollution mitigation. Predictive models demonstrate significant potential for forecasting air quality, with performance metrics such as RMSE, MAE, and MAPE underscoring their accuracy. The findings advocate for integrated monitoring and strategic urban planning to improve air quality, emphasizing the role of machine learning in environmental science.