2023 No https://doi.org/10.1016/j.enbuild.2023.112787 Yang, Guangfei Asia Exposure Study Modeling Students Elementary (K-5) Temperature Humidity Particulate Matter (PM) Indoor PM2.5 Air pollution in classrooms Machine learning Data-driven Indoor PM2.5 was monitored for 12 months in 10 classrooms using data-driven models. Partial derivative analysis revealed seasonal dynamics and influential contributors. Findings support sensor-based strategies to improve classroom air quality.