Comparative Evaluation of Machine Learning Models for Residential PM1 Prediction in Zagreb (Croatia): Identifying Key Predictors and Indoor/Outdoor Dynamics
Authors:
- Marija Jelena Lovrić ŠtefičekMŠMarija Jelena Lovrić Štefiček
- Silvije DavilaSDSilvije Davila
- Gordana PehnecGPGordana Pehnec
- Ivan BešlićIBIvan Bešlić
- Željka Ujević AndrijićŽAŽeljka Ujević Andrijić
- Ivana BanićIBIvana Banić
- Mirjana TurkaljMTMirjana Turkalj
- Mario LovrićMario Lovrić
- Luka KazenskyLKLuka Kazensky
- Goran GajskiGoran Gajski
Open Access
Abstract
Indoor exposure to particulate matter (PM) is increasingly recognized as a major contributor to respiratory and cardiovascular risk, yet the relative contributions of outdoor pollution, building characteristics, and occupant behavior remain poorly resolved. PM1 (aerodynamic diameter < 1 μm) warrants focus due to its higher alveolar deposition. “Evidence driven indoor air quality improvement” (EDIAQI) project aims to enhance indoor air quality guidelines and increase awareness by providing accessible data on exposure, pollution sources, and related risk factors. As part of the Zagreb pilot within the project, 103 paired indoor/outdoor PM1 samples were analyzed. Seasonal analysis revealed substantial wintertime outdoor PM1 spikes, while indoor medians remained stable. Chemometric analysis identified factors such as dwelling size, outdoor pollution, resuspension, building age/heating type, and urban context. Among the tested models, the validated gradient-boosted regressor (GBR) achieved the strongest performance, explaining ~65% variance in indoor PM1 (test R2 ≈ 0.65). Explainable machine learning analysis (SHAP) identified outdoor PM1 levels, infiltration, and resuspension as the most influential predictors. Findings underscore wintertime outdoor emissions (e.g., residential heating and traffic) and dwelling-related and behavioral factors as key drivers, with the machine learning–environmental data integration enabling targeted residential IAQ management: optimized ventilation protocols, resuspension mitigation via behavior, and infiltration reduction through retrofits.Keywords
Funding Information
- research project “Evidence driven indoor air quality improvement” (101057497)
- European Regional Development Fund (KK.01.1.1.02.0007)
- Institute for Medical Research and Occupational Health and the European Union—Next Generation EU (533-03-23-0006)
This publication has 81 references indexed in Scilit:
- Machine Learning for Indoor Air Quality Assessment: A Systematic Review and AnalysisEnvironmental Modeling & Assessment, 2024
- Evidence driven indoor air quality improvement: An innovative and interdisciplinary approach to improving indoor air qualityBioFactors, 2024
- Impact of indoor air pollution on DNA damage and chromosome stability: a systematic reviewArchives of Toxicology, 2024
- Supervised feature selection using principal component analysisKnowledge and Information Systems, 2023
- Developing PM2.5 and PM10 prediction models on a national and regional scale using open-source remote sensing dataEnvironmental Monitoring and Assessment, 2023
- Machine learning algorithms to forecast air quality: a surveyArtificial Intelligence Review, 2023
- Effect of particulate matter (PM2.5 and PM10) on health indicators: climate change scenarios in a Brazilian metropolisEnvironmental Geochemistry and Health, 2022
- Household air pollution from cooking and heating and its impacts on blood pressure in residents living in rural cave dwellings in Loess Plateau of ChinaEnvironmental Science and Pollution Research, 2020
- Polycyclic aromatic hydrocarbons in PM10, PM2.5 and PM1 particle fractions in an urban areaAir Quality, Atmosphere & Health, 2018
- Review of relationship between indoor and outdoor particles: I/O ratio, infiltration factor and penetration factorAtmospheric Environment, 2011