Work

Generative physics-informed machine learning for modeling indoor air quality and its impact on student health and performance

Year: 2025

Type: article

Abstract: This study presents a novel hybrid framework combining experimental testing, statistical analysis, physics-based modeling, and interpretable machine learning (ML) to assess indoor air quality (IAQ) and its impact on student health and academic performance in healthcare and laboratory environments. A comparative feature importance analysis using Spearman, Kendall methods, and SHAP values identified particulate matter (PM 10 ) as the most critical pollutant, with location, noise, and thermal factors playing secondary roles. Empirical assessments revealed concerning IAQ variations–Lab 187E (Biological Prep) exhibited elevated PM 2.5 (12.4 μg/m 3 ), while Lab 290D (Chemical Storage) showed high TVOC (312 ppb) and Virus Index (3). At Bilborough Medical Centre, CO 2 peaked at 2819 ppm (exceeding NHS limits), and TVOC reached 2618 ppb, highlighting ventilation deficiencies. The PDE-XGBoost model outperformed PDE-RF (MAE: 0.4548 vs. 0.6652; R 2 : 0.989 vs. 0.976), demonstrating superior predictive accuracy for IAQ parameters. Partial Dependence Plots (PDPs) revealed TVOC and noise as dominant linear predictors of PM 2 .5, while temperature exhibited a nonlinear effect. SHAP analysis underscored PM 10 direct physiological impact and location's indirect role in pollutant dispersion. The integration of statistical and ML methods provided actionable insights: PM 10 mitigation, spatial optimization, and VOC control are critical for healthier environments. • Hybrid IAQ framework fuses experiments, physics models, statistics, and ML. • PM10 is the key pollutant; location, noise, thermal factors are secondary. • Clinic CO 2 2819 ppm and TVOC 2618 ppb signal ventilation shortfalls. • PDE-XGBoost outperformed PDE-RF (MAE 0.455 vs 0.665; R 2 0.989 vs 0.976). • PDPs/SHAP guide actions: reduce PM10, optimize space, control VOC sources.

Source: Journal of Environmental Management

DOI: 10.1016/j.jenvman.2025.127676

Authors: Adekunle Dosumu, Seyed Hamed Godasiaei, Obuks A. Ejohwomu, Saeed Reza Mohandes

Institutions: Nottingham Trent University, Xi'an Jiaotong University, University of Johannesburg, University of Manchester

Language: English

Topic: Indoor Air Quality and Microbial Exposure

Subfield: Health, Toxicology and Mutagenesis

Field: Environmental Science

Domain: Physical Sciences

SDG: Good health and well-being

Open Access status: hybrid

Estimated APC paid: $3,750