On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models

Air pollution, particularly fine (PM<sub>2.5</sub>) and coarse (PM<sub>10</sub>) particulate matter, poses significant risks to public health and environmental sustainability. This study aims to develop robust predictive and forecasting models for hourly PM concentrations in...

Full description

Saved in:
Bibliographic Details
Main Authors: Youness El Mghouchi, Mihaela Tinca Udristioiu
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/15/8254
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849405716989739008
author Youness El Mghouchi
Mihaela Tinca Udristioiu
author_facet Youness El Mghouchi
Mihaela Tinca Udristioiu
author_sort Youness El Mghouchi
collection DOAJ
description Air pollution, particularly fine (PM<sub>2.5</sub>) and coarse (PM<sub>10</sub>) particulate matter, poses significant risks to public health and environmental sustainability. This study aims to develop robust predictive and forecasting models for hourly PM concentrations in Craiova, Romania, using advanced hybrid Artificial Intelligence (AI) approaches. A five-year dataset (2020–2024), comprising 20 meteorological and pollution-related variables recorded by four air quality monitoring stations, was analyzed. The methodology consists of three main phases: (i) data preprocessing, including anomaly detection and missing value handling; (ii) exploratory analysis to identify trends and correlations between PM concentrations (PMs) and predictor variables; and (iii) model development using 23 machine learning and deep learning algorithms, enhanced by 50 feature selection techniques. A deep Nonlinear AutoRegressive Moving Average with eXogenous inputs (Deep-NARMAX) model was employed for multi-step-ahead forecasting. The best-performing models achieved R<sup>2</sup> values of 0.85 for PM<sub>2.5</sub> and 0.89 for PM<sub>10</sub>, with low RMSE and MAPE scores, demonstrating high accuracy and generalizability. The GEO-based feature selection method effectively identified the most relevant predictors, while the Deep-NARMAX model captured temporal dynamics for accurate forecasting. These results highlight the potential of hybrid AI models for air quality management and provide a scalable framework for urban pollution monitoring, predicting, and forecasting.
format Article
id doaj-art-ff9c802c68d543b6b567c3d487bf33e8
institution Kabale University
issn 2076-3417
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-ff9c802c68d543b6b567c3d487bf33e82025-08-20T03:36:35ZengMDPI AGApplied Sciences2076-34172025-07-011515825410.3390/app15158254On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based ModelsYouness El Mghouchi0Mihaela Tinca Udristioiu1Department of Energetics, École Nationale Supérieure d’Arts et Métiers—ENSAM, Moulay Ismail University, Meknes 15290, MoroccoDepartment of Physics, Faculty of Sciences, University of Craiova, 13 A.I. Cuza Street, 200585 Craiova, RomaniaAir pollution, particularly fine (PM<sub>2.5</sub>) and coarse (PM<sub>10</sub>) particulate matter, poses significant risks to public health and environmental sustainability. This study aims to develop robust predictive and forecasting models for hourly PM concentrations in Craiova, Romania, using advanced hybrid Artificial Intelligence (AI) approaches. A five-year dataset (2020–2024), comprising 20 meteorological and pollution-related variables recorded by four air quality monitoring stations, was analyzed. The methodology consists of three main phases: (i) data preprocessing, including anomaly detection and missing value handling; (ii) exploratory analysis to identify trends and correlations between PM concentrations (PMs) and predictor variables; and (iii) model development using 23 machine learning and deep learning algorithms, enhanced by 50 feature selection techniques. A deep Nonlinear AutoRegressive Moving Average with eXogenous inputs (Deep-NARMAX) model was employed for multi-step-ahead forecasting. The best-performing models achieved R<sup>2</sup> values of 0.85 for PM<sub>2.5</sub> and 0.89 for PM<sub>10</sub>, with low RMSE and MAPE scores, demonstrating high accuracy and generalizability. The GEO-based feature selection method effectively identified the most relevant predictors, while the Deep-NARMAX model captured temporal dynamics for accurate forecasting. These results highlight the potential of hybrid AI models for air quality management and provide a scalable framework for urban pollution monitoring, predicting, and forecasting.https://www.mdpi.com/2076-3417/15/15/8254air pollutionPMshybrid AImodelspredictionforecasting
spellingShingle Youness El Mghouchi
Mihaela Tinca Udristioiu
On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
Applied Sciences
air pollution
PMs
hybrid AI
models
prediction
forecasting
title On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
title_full On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
title_fullStr On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
title_full_unstemmed On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
title_short On the Prediction and Forecasting of PMs and Air Pollution: An Application of Deep Hybrid AI-Based Models
title_sort on the prediction and forecasting of pms and air pollution an application of deep hybrid ai based models
topic air pollution
PMs
hybrid AI
models
prediction
forecasting
url https://www.mdpi.com/2076-3417/15/15/8254
work_keys_str_mv AT younesselmghouchi onthepredictionandforecastingofpmsandairpollutionanapplicationofdeephybridaibasedmodels
AT mihaelatincaudristioiu onthepredictionandforecastingofpmsandairpollutionanapplicationofdeephybridaibasedmodels