Deep Hybrid AI Models Applied to Predict, Model, and Forecast the Next Upcoming Periods of Ozone in Craiova City
Source abstract
Abstract
Artificial intelligence (AI) plays an important role in analyzing air quality, providing new insights that enable informed environmental policy decisions at the local level based on air pollution modeling and forecasting. The aim of this study is to analyze various hybrid AI methods to predict, model, and anticipate hourly ground-level ozone concentrations. Ground-level ozone concentrations impact human health and the environment. The data used in this study was downloaded from the website of the Romanian Agency for Environmental Protection and spans five years (2020–2024). The dataset comprises two categories of data: (i) seven meteorological parameters, including temperature (T), relative humidity, precipitation, air pressure, solar brightness, wind direction, and velocity; (ii) twenty air pollutants, including two types of particulate matter, carbon monoxide, sulfur dioxide, ground-level ozone, three types of nitrogen oxide, ammonia, six volatile organic compounds, and five toxic elements. The study follows a six-stage approach: (1) data preprocessing is conducted to identify and address anomalies, outliers, and missing values, while ozone trends are analyzed; (2) correlations between ozone concentrations and other variables are examined, considering only non-missing values; (3) data splitting is carried out in training and testing sets; (4) a total of 27 hybrid AI-based algorithms are applied to determine the optimal predictive model for ozone concentration based on related variables; (5) fifty feature selection methods are applied to find the most relevant predictors for predicting ozone concentration; (6) a novel deep NARMAX model is employed to model and anticipate hourly ozone levels in Craiova. Using a set of statistical metrics, the results of the models are assessed. This research provides a novel perspective on the robustness of the predictive performance of the proposed model.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Air Quality Monitoring and Forecasting · Air Quality and Health Impacts · COVID-19 impact on air quality
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Physics, Faculty of Sciences, University of Craiova, 13 A.I. Cuza Street, 200585 Craiova, Romania
Department of Energetics, École Nationale Supérieure d’Arts et Métiers (ENSAM), Moulay Ismail University, Meknes 15290, Morocco
References (100+)
Show source-provided references
1. Mohamed, A.-M.O., Paleologos, E.K., and Howari, F.M. (2021). Chapter 12-Health economics of air pollution. Pollution Assessment for Sustainable Practices in Applied Sciences and Engineering, Butterworth-Heineman.
2. Analyzing the interconnected dynamics of domestic biofuel burning in India: Unravelling VOC emissions, surface-ozone formation, diagnostic ratios, and source identification10.1039/D4SU00030G ↗
3. Manisalidis, I., Stavropoulou, E., Stavropoulos, A., and Bezirtzoglou, E. (2020). Environmental and Health Impacts of Air Pollution: A Review. Front. Public Health, 8.10.3389/fpubh.2020.00014 ↗
4. Prediction, modelling, and forecasting of PM and AQI using hybrid machine learning10.1016/j.jclepro.2023.138496 ↗
5. Forecasting ground-level ozone and fine particulate matter concentrations at Craiova city using a meta-hybrid deep learning model10.1016/j.uclim.2024.102099 ↗
6. El Mghouchi, Y., Udristioiu, M.T., and Yildizhan, H. (2024). Multivariable Air-Quality Prediction and Modelling via Hybrid Machine Learning: A Case Study for Craiova, Romania. Sensors, 24.10.3390/s24051532 ↗
7. Cardiovascular function and ozone exposure: The Multicenter Ozone Study in oldEr Subjects (MOSES)10.1016/j.envint.2018.06.014 ↗
8. Effects of ozone on agriculture, forests and grasslands
9. Forecasting ground-level ozone concentration levels using machine learning10.1016/j.resconrec.2022.106380 ↗
10. Forecasting of fine particulate matter based on LSTM and optimization algorithm10.1016/j.jclepro.2023.139233 ↗
11. Artificial intelligence: Machine learning for chemical sciences10.1007/s12039-021-01995-2 ↗
12. Prediction of ozone levels using a Hidden Markov Model (HMM) with Gamma distribution10.1016/j.atmosenv.2012.08.008 ↗
13. Autoregressive models in environmental forecasting time series: A theoretical and application review10.1007/s11356-023-25148-9 ↗
14. Gao, W., Xiao, T., Zou, L., Li, H., and Gu, S. (2024). Analysis and Prediction of Atmospheric Environmental Quality Based on the Autoregressive Integrated Moving Average Model (ARIMA Model) in Hunan Province, China. Sustainability, 16.10.3390/su16198471 ↗
15. IoT based improved air quality index prediction using hybrid FA-ANN-ARMA model
16. Forecasting the Air Quality Index by Utilizing Several Meteorological Factors Using the ARIMAX Method (Case Study: Central Jakarta City)10.35870/jtik.v8i3.2012 ↗
17. Time Series Analysis and Forecasting of Air Pollution Particulate Matter (PM2.5): An SARIMA and Factor Analysis Approach10.1109/ACCESS.2021.3060744 ↗
18. A Hybrid Air Quality Prediction Method Based on VAR and Random Forest10.4236/jcc.2025.132009 ↗
19. Soybean and Soybean Oil Price Forecasting through the Nonlinear Autoregressive Neural Network (NARNN) and NARNN with Exogenous Inputs (NARNN–X)
20. Air pollution prediction by using an artificial neural network model10.1007/s10098-019-01709-w ↗
21. Naveen, S., Upamanyu, M.S., Chakki, K.M.C., and Hariprasad, P. (2023, January 7–8). Air Quality Prediction Based on Decision Tree Using Machine Learning. Proceedings of the 2023 International Conference on Smart Systems for Applications in Electrical Sciences (ICSSES), Tumakuru, India.10.1109/ICSSES58299.2023.10200535 ↗
22. Spatio-temporal prediction and factor identification of urban air quality using support vector machine10.1016/j.uclim.2021.101055 ↗
23. Baran, B. (2019, January 21–22). Prediction of Air Quality Index by Extreme Learning Machines. Proceedings of the 2019 International Artificial Intelligence and Data Processing Symposium (IDAP), Malatya, Turkey.10.1109/IDAP.2019.8875910 ↗
24. Shyamala, K., and Sujatha, R. (2024). Modified Extreme Gradient Boosting Algorithm for Prediction of Air Pollutants in Various Peak Hours. Advancements in Smart Computing and Information Security, Springer.10.1007/978-3-031-58604-0_9 ↗
25. Reddy, P.D., and Parvathy, L.R. (2022, January 20–22). Prediction Analysis using Random Forest Algorithms to Forecast the Air Pollution Level in a Particular Location. Proceedings of the 2022 3rd International Conference on Smart Electronics and Communication (ICOSEC), Trichy, India.10.1109/ICOSEC54921.2022.9952138 ↗
26. Link, S., and Prade, H. (2010). Bagging Decision Trees on Data Sets with Classification Noise. Foundations of Information and Knowledge Systems, Springer.10.1007/978-3-642-11829-6 ↗
27. Collinearity in generalized linear regression10.1080/03610929308831126 ↗
28. Modeling of subway indoor air quality using Gaussian process regression10.1016/j.jhazmat.2018.07.034 ↗
29. Sonu, S.B., and Suyampulingam, A. (2021, January 27–28). Linear Regression Based Air Quality Data Analysis and Prediction using Python. Proceedings of the 2021 IEEE Madras Section Conference (MASCON), Chennai, India.10.1109/MASCON51689.2021.9563432 ↗
30. Generalized additive models: Building evidence of air pollution, climate change and human health10.1016/j.envint.2019.104987 ↗
31. Schölkopf, B., Luo, Z., and Vovk, V. (2013). Kernel Ridge Regression. Empirical Inference: Festschrift in Honor of Vladimir, N. Vapnik, Springer.10.1007/978-3-642-41136-6 ↗
32. Comparative Analysis Study for Air Quality Prediction in Smart Cities Using Regression Techniques10.1109/ACCESS.2023.3323447 ↗
33. Spatial assessment of PM10 hotspots using Random Forest, K-Nearest Neighbour and Naïve Bayes10.1016/j.apr.2021.101202 ↗
34. Su, Y. (2020, January 23–25). Prediction of air quality based on Gradient Boosting Machine Method. Proceedings of the 2020 International Conference on Big Data and Informatization Education (ICBDIE), Zhangjiajie, China.10.1109/ICBDIE50010.2020.00099 ↗
35. Bayesian network reasoning and machine learning with multiple data features: Air pollution risk monitoring and early warning10.1007/s11069-021-04504-3 ↗
36. DeepAirNet: Applying Recurrent Networks for Air Quality Prediction10.1016/j.procs.2018.05.068 ↗
37. Air quality modelling using long short-term memory (LSTM) over NCT-Delhi, India10.1007/s11869-019-00696-7 ↗
38. Air Quality Prediction in Visakhapatnam with LSTM based Recurrent Neural Networks
39. Chakma, A., Vizena, B., Cao, T., Lin, J., and Zhang, J. (2017, January 17–20). Image-based air quality analysis using deep convolutional neural network. Proceedings of the 2017 IEEE International Conference on Image Processing (ICIP), Beijing, China.10.1109/ICIP.2017.8297023 ↗
40. A pruned feed-forward neural network (pruned-FNN) approach to measure air pollution exposure10.1007/s10661-023-11814-5 ↗
41. A systematic literature review of deep learning neural network for time series air quality forecasting10.1007/s11356-021-17442-1 ↗
42. RCL-Learning: ResNet and convolutional long short-term memory-based spatiotemporal air pollutant concentration prediction model10.1016/j.eswa.2022.118017 ↗
43. Identifying Contributors to PM2.5 Simulation Biases of Chemical Transport Model Using Fully Connected Neural Networks10.1029/2021MS002898 ↗
44. Reinforcement Learning for Improving the Accuracy of PM2.5 Pollution Forecast Under the Neural Network Framework10.1109/ACCESS.2019.2932413 ↗
45. Temporal convolutional denoising autoencoder network for air pollution prediction with missing values10.1016/j.uclim.2021.100872 ↗
46. Predictive maintenance in pharmaceutical manufacturing lines using deep transformers10.1016/j.procs.2023.03.073 ↗
47. Dong, J., Zhang, Y., and Hu, J. (2024). Short-term air quality prediction based on EMD-transformer-BiLSTM. Sci. Rep., 14.10.1038/s41598-024-67626-1 ↗
48. Tian, J., Liang, Y., Xu, R., Chen, P., Guo, C., Zhou, A., Pan, L., Rao, Z., and Yang, B. (2025). Air Quality Prediction with Physics-Informed Dual Neural ODEs in Open Systems. arXiv.
49. Air pollutant prediction based on ARIMA-WOA-LSTM model10.1016/j.apr.2023.101761 ↗
50. Unsupervised outlier detection for time-series data of indoor air quality using LSTM autoencoder with ensemble method10.1186/s40537-023-00746-z ↗
51. Novel climate classification based on the information of solar radiation intensity: An application to the climatic zoning of Morocco10.1016/j.enconman.2021.114770 ↗
52. Predictive maintenance in Industry 4.0: A survey of planning models and machine learning techniques10.7717/peerj-cs.2016 ↗
53. Analysis of the particulate matter long term emissions in Romania by sectors of activities
54. Moldovan, C.S., Mateescu, M.D., Sbirna, L.S., Ionescu, C., and Sbirna, S. (2011). Study regarding concentration of main air pollutants in Craiova (Romania)–Comparison between the first half of 2011 and the similar period of 2010. SESAM, INSEMEX.
55. Buzatu, G.D., and Dodocioiu, A.M. (2022). Air quality study in Craiova municipality based on data provided by uRADm independent sensor network. Ann. Univ. Craiova Biol. Hortic. Food Prod. Process. Technol. Environ. Eng., 27.10.52846/bihpt.v27i63.17 ↗
56. Velea, L., Udriștioiu, M.T., Puiu, S., Motișan, R., and Amarie, D. (2023). A Community-Based Sensor Network for Monitoring the Air Quality in Urban Romania. Atmosphere, 14.10.3390/atmos14050840 ↗
57. First results given by the independent air pollution monitoring network from Craiova city Romania10.1063/5.0152017 ↗
58. Pekdogan, T., Udriștioiu, M.T., Yildizhan, H., and Ameen, A. (2024). From Local Issues to Global Impacts: Evidence of Air Pollution for Romania and Turkey. Sensors, 24.10.3390/s24041320 ↗
59. Yildizhan, H., Udriștioiu, M.T., Pekdogan, T., and Ameen, A. (2024). Observational study of ground-level ozone and climatic factors in Craiova, Romania, based on one-year high-resolution data. Sci. Rep., 14.10.1038/s41598-024-77989-0 ↗
60. Using wavelet–feedforward neural networks to improve air pollution forecasting in urban environments10.1007/s10661-015-4697-x ↗
61. Examining effects of air pollution on photovoltaic systems via interpretable random forest model10.1016/j.renene.2024.121066 ↗
62. Characteristics of extreme temperatures relevant for agriculture in the near future (2021–2040) in Romania
63. Brâncuș, M.M., Burada, C., and Mănescu, C. (2023, January 3–7). The impact of late and early snowfall on urban areas in southwestern Romania. Proceedings of the Climate, Water and Society: Changes and Challenges, 36th Conference of the International Association of Climatology, Bucharest, Romania.
64. Assessment of urban heat Island for Craiova from satellite-based LST10.1063/1.5017443 ↗
65. Assessing the performance of solar radiation computing models and model selection procedures10.1016/j.jastp.2013.09.004 ↗
66. Induction of decision trees10.1023/A:1022643204877 ↗
67. (2025, March 15). Variance Threshold as Early Screening to Boruta Feature Selection for Intrusion Detection System. Available online: https://ieeexplore.ieee.org/abstract/document/9608852.
68. Feature Selection Using a Neural Network with Group Lasso Regularization and Controlled Redundancy10.1109/TNNLS.2020.2980383 ↗
69. Supervised feature selection using principal component analysis10.1007/s10115-023-01993-5 ↗
70. Feature Selection Using Particle Swarm Optimization in Intrusion Detection
71. Regression modeling for enterprise electricity consumption: A comparison of recurrent neural network and its variants10.1016/j.ijepes.2020.106612 ↗
72. Support-Vector Networks10.1023/A:1022627411411 ↗
73. Enhanced performance Gaussian process regression for probabilistic short-term solar output forecast10.1016/j.ijepes.2021.106916 ↗
74. Mousavirad, S.J., and Ebrahimpour-Komleh, H. (November, January 31). Feature selection using modified imperialist competitive algorithm. Proceedings of the ICCKE 2013, Mashhad, Iran.10.1109/ICCKE.2013.6682833 ↗
75. Hybrid filter-wrapper feature selection using whale optimization algorithm: A multi-objective approach10.1016/j.eswa.2021.115312 ↗
76. Symbolic Regression Methods for Reinforcement Learning10.1109/ACCESS.2021.3119000 ↗
77. A hybrid Harris Hawks optimization algorithm with simulated annealing for feature selection10.1007/s10462-020-09860-3 ↗
78. Comparing regression, naive Bayes, and random forest methods in the prediction of individual survival to second lactation in Holstein cattle10.3168/jds.2019-16295 ↗
79. Feature selection for high dimensional imbalanced class data using harmony search10.1016/j.engappai.2016.10.008 ↗
80. A wrapper approach for feature selection based on Bat Algorithm and Optimum-Path Forest10.1016/j.eswa.2013.09.023 ↗
81. Extreme learning machine: Theory and applications10.1016/j.neucom.2005.12.126 ↗
82. A Review on Linear Regression Comprehensive in Machine Learning10.38094/jastt1457 ↗
83. Hastie, T.J. (1992). Generalized Additive Models. Statistical Models in S, Routledge.
84. The raven roosting optimisation algorithm10.1007/s00500-014-1520-5 ↗
85. Current Studies and Applications of Shuffled Frog Leaping Algorithm: A Review10.1007/s11831-021-09707-2 ↗
86. A New Hybrid Seagull Optimization Algorithm for Feature Selection10.1109/ACCESS.2019.2909945 ↗
87. Improved salp swarm algorithm for feature selection10.1016/j.jksuci.2018.06.003 ↗
88. An efficient instance selection algorithm for k nearest neighbor regression10.1016/j.neucom.2017.04.018 ↗
89. Quantification of an Adverse Outcome Pathway Network by Bayesian Regression and Bayesian Network Modeling10.1002/ieam.4348 ↗
90. Suma, V., Fernando, X., Du, K.L., and Wang, H. (2022). Feature Selection Using Modified Sine Cosine Algorithm with COVID-19 Dataset. Evolutionary Computing and Mobile Sustainable Networks, Springer.10.1007/978-981-16-9605-3 ↗
91. Feature selection via Lèvy Antlion optimization10.1007/s10044-018-0695-2 ↗
92. A Review of Glowworm Swarm Optimization Meta-Heuristic Swarm Intelligence and its Fusion in Various Applications
93. Hafez, A.I., Hassanien, A.E., Zawbaa, H.M., and Emary, E. (2015, January 29–30). Hybrid Monkey Algorithm with Krill Herd Algorithm optimization for feature selection. Proceedings of the 2015 11th International Computer Engineering Conference (ICENCO), Cairo, Egypt.10.1109/ICENCO.2015.7416361 ↗
94. Optimal feature selection for SAR image classification using biogeography-based optimization (BBO), artificial bee colony (ABC) and support vector machine (SVM): A combined approach of optimization and machine learning10.1007/s10596-020-10030-1 ↗
95. Using data complexity measures and an evolutionary cultural algorithm for gene selection in microarray data10.1016/j.socl.2020.100007 ↗
96. A local opposition-learning golden-sine grey wolf optimization algorithm for feature selection in data classification10.1016/j.asoc.2023.110319 ↗
97. Natekin, A., and Knoll, A. (2013). Gradient boosting machines, a tutorial. Front. Neurorobot., 7.10.3389/fnbot.2013.00021 ↗
98. Nadimi-Shahraki, M.H., Zamani, H., and Mirjalili, S. (2022). Enhanced whale optimization algorithm for medical feature selection: A COVID-19 case study. Comput. Biol. Med., 148.10.1016/j.compbiomed.2022.105858 ↗
99. Feature selection for chemical process fault diagnosis by artificial immune systems10.1016/j.cjche.2017.09.023 ↗
100. Eagle Strategy in Nature-Inspired Optimization: Theory, Analysis, Applications, and Comparative Study10.1007/s11831-023-10014-1 ↗