Back to Search
Start Over
Construction and evaluation of hourly average indoor PM 2.5 concentration prediction models based on multiple types of places.
- Source :
-
Frontiers in public health [Front Public Health] 2023 Aug 10; Vol. 11, pp. 1213453. Date of Electronic Publication: 2023 Aug 10 (Print Publication: 2023). - Publication Year :
- 2023
-
Abstract
- Background: People usually spend most of their time indoors, so indoor fine particulate matter (PM <subscript>2.5</subscript> ) concentrations are crucial for refining individual PM <subscript>2.5</subscript> exposure evaluation. The development of indoor PM <subscript>2.5</subscript> concentration prediction models is essential for the health risk assessment of PM <subscript>2.5</subscript> in epidemiological studies involving large populations.<br />Methods: In this study, based on the monitoring data of multiple types of places, the classical multiple linear regression (MLR) method and random forest regression (RFR) algorithm of machine learning were used to develop hourly average indoor PM <subscript>2.5</subscript> concentration prediction models. Indoor PM <subscript>2.5</subscript> concentration data, which included 11,712 records from five types of places, were obtained by on-site monitoring. Moreover, the potential predictor variable data were derived from outdoor monitoring stations and meteorological databases. A ten-fold cross-validation was conducted to examine the performance of all proposed models.<br />Results: The final predictor variables incorporated in the MLR model were outdoor PM <subscript>2.5</subscript> concentration, type of place, season, wind direction, surface wind speed, hour, precipitation, air pressure, and relative humidity. The ten-fold cross-validation results indicated that both models constructed had good predictive performance, with the determination coefficients (R <superscript>2</superscript> ) of RFR and MLR were 72.20 and 60.35%, respectively. Generally, the RFR model had better predictive performance than the MLR model (RFR model developed using the same predictor variables as the MLR model, R <superscript>2</superscript> = 71.86%). In terms of predictors, the importance results of predictor variables for both types of models suggested that outdoor PM <subscript>2.5</subscript> concentration, type of place, season, hour, wind direction, and surface wind speed were the most important predictor variables.<br />Conclusion: In this research, hourly average indoor PM <subscript>2.5</subscript> concentration prediction models based on multiple types of places were developed for the first time. Both the MLR and RFR models based on easily accessible indicators displayed promising predictive performance, in which the machine learning domain RFR model outperformed the classical MLR model, and this result suggests the potential application of RFR algorithms for indoor air pollutant concentration prediction.<br />Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.<br /> (Copyright © 2023 Shi, Du, Zhang, Han, Chen, Wang, Liu, Zhang, Dong and Sui.)
- Subjects :
- Humans
Seasons
Algorithms
Databases, Factual
Particulate Matter
Air Pollutants
Subjects
Details
- Language :
- English
- ISSN :
- 2296-2565
- Volume :
- 11
- Database :
- MEDLINE
- Journal :
- Frontiers in public health
- Publication Type :
- Academic Journal
- Accession number :
- 37637795
- Full Text :
- https://doi.org/10.3389/fpubh.2023.1213453