Comparison of Machine Learning and Deep Learning Methods for Modeling Ozone Concentrations

Abstract

Although air pollution is an important problem for today, reasons such as industrialization, forest fires, exhaust gases, poor quality fuel use confront us with a serious problem that will threaten future generations. Among these reasons, intensive industrialization is one of the most critical factors that play a role in air pollution. Regional industrial development affects air quality in cities. While the amount of some pollutants decreases with the development of the industry, there is an increase in ozone levels. In the coming years, it becomes inevitable to predict air pollution in order not to feel the problems that air pollution will cause more, to manage air quality, and to take precautions against risks. In this study, machine learning and deep learning methods based on time series were applied to predict hourly ozone levels between 2018 and 2021 for the provinces of Kocaeli and Sakarya, where the industry is developed, and Çanakkale, where the industry is not developed much. The applied models were compared using Mean Absolute Error (MAE), Relative Absolute Error (RAE), and R-square (R2) metrics, and it was aimed to determine the most effective method.

Keywords

• Machine learning
• Deep learning
• Extreme Gradient Boosting (Xgboost)
• Artificial Neural Network (ANN)
• Long-Short Term Memory (LSTM)
• Time Series

Ozon Konsantrasyonlarını Modellemek için Makine Öğrenmesi ve Derin Öğrenme Yöntemlerinin Karşılaştırılması

Abstract

Hava kirliliği günümüz için önemli bir problem olmakla birlikte sanayileşme, orman yangınları, egzoz gazları, kalitesiz yakıt kullanımı gibi sebepler gelecek nesilleri de tehdit edecek ciddi bir problem ile bizleri yüzleştirmektedir. Bu sebepler içerisinde ise yoğun sanayileşme hava kirliliğinde rol oynayan en önemli faktörlerden birisidir. Bölgesel sanayi gelişimi şehirlerde hava kalitesini etkilemektedir. Sanayinin gelişmesi ile birlikte bazı kirleticilerin miktarı azalmakta iken, ozon seviyelerinde artış yaşanmaktadır. Önümüzdeki yıllarda hava kirliliğini neden olacağı problemleri daha fazla hissetmemek, hava kalitesini yönetmek ve risklere karşı önlem almak için hava kirliliğinin tahmin edilmesi kaçınılmaz hale gelmektedir. Bu çalışmada sanayinin gelişmiş olduğu Kocaeli ve Sakarya illeri ile sanayinin çok fazla gelişmediği Çanakkale illeri için 2018-2021 arası saatlik ozon seviyelerini tahmin etmek amacıyla zaman serilerine dayalı makine öğrenmesi ve derin öğrenme yöntemleri uygulanmıştır. Uygulanan modeller Ortalama Mutlak Hata (MAE), Bağıl Mutlak Hata (RAE) ve R-kare (R2) metrikleri kullanılarak karşılaştırılmış ve en etkin yöntemin belirlenmesi amaçlanmıştır.

Keywords

• Makine Öğrenmesi
• Derin Öğrenme
• Aşırı Gradyan Arttırma (Xgboost)
• Yapay Sinir Ağları (YSA)
• Uzun Kısa Süreli Bellek (LSTM)
• Zaman Serileri

References

1. Adnane, A., Leghrib, R., Chaoufi, J., & Chirmata, A., 2020. The Use of a Recurrent Neural Network for Forecasting Ozone Concentrations in the City of Agadir (Morocco). Journal of Atomic, Molecular, Condensed Matter and Nano Physics, 7(3), 197-206.
2. Alghieth, M., Alawaji, R., Saleh, S. H., Alh, S., 2021. Air Pollution Forecasting Using Deep Learning. International Journal of Online & Biomedical Engineering, 17(14).
3. Alipio, M. M., 2020. Do latitude and ozone concentration predict Covid-2019 cases in 34 countries?. medRxiv.
4. Allu, S. K., Srinivasan, S., Maddala, R. K., Reddy, A., Anupoju, G. R., 2020. Seasonal ground level ozone prediction using multiple linear regression (MLR) model. Modeling Earth Systems and Environment, 6, 1981-1989.
5. Bekesiene, S., Meidute-Kavaliauskiene, I., Vasiliauskiene, V., 2021. Accurate prediction of concentration changes in ozone as an air pollutant by multiple linear regression and artificial neural networks. Mathematics, 9(4), 356.
6. Bilgin, G., 2021. Investigation of The Risk of Diabetes in Early Period using Machine Learning. Journal of Intelligent Systems: Theory and Applications, 4(1), 55-64.
7. Chattopadhyay, G., Midya, S. K., Chattopadhyay, S., 2019. MLP based predictive model for surface ozone concentration over an urban area in the Gangetic West Bengal during pre-monsoon season. Journal of Atmospheric and Solar-Terrestrial Physics, 184, 57-62.
8. Chelani, A. B., 2010. Prediction of daily maximum ground ozone concentration using support vector machine. Environmental monitoring and assessment, 162(1), 169-176.

9. Çağıl, G., Yıldırım, B., 2020. Detection of an Assembly Part with Deep Learning and Image Processing. Journal of Intelligent Systems: Theory and Applications, 3(2), 31-37.
10. Darendeli, B. N., Yılmaz, A., 2021. Convolutional Neural Network Approach to Predict Tumor Samples Using Gene Expression Data. Journal of Intelligent Systems: Theory and Applications, 4(2), 136-141.
11. Ding, J., Liu, M., Ma, Z., Liu, R., Bi, J., 2020. Spatial and temporal trends in the mortality burden of ozone pollution in China: 2005-2017. ISEE Conference Abstracts, 24-27 August 2020.
12. Ekinci, E., İlhan Omurca, S., Özbay, B., 2021. Comparative assessment of modeling deep learning networks for modeling ground-level ozone concentrations of pandemic lock-down period. Ecological Modelling, 457, 1-11.
13. Ekinci, E., İlhan Omurca, S., Sevim, S., 2020. Improve Offensive Language Detection with Ensemble Classifiers. International Journal of Intelligent Systems and Applications in Engineering, 8(2), 109-115.
14. Eslami, E., Choi, Y., Lops, Y., Sayeed, A., 2020. A real-time hourly ozone prediction system using deep convolutional neural network. Neural Computing and Applications, 32(13), 8783-8797.
15. Garip Batık, Z., Büyükbıçakçı, E., 2016. Klasik Enterpolasyon Yöntemleri ve Yapay Sinir Ağı Yaklaşımları ile Matematiksel Denklemlerin Karşılaştırılmalı Çözümü İçin Arayüz Tasarımı, 4th International Symposium on Innovative Technologies in Engineering and Science, 3-5 November 2016, Antalya, Turkey, pp. 1379-1383.
16. Kleinert, F., Leufen, L. H., Lupascu, A., Butler, T., Schultz, M. G., 2021. Representing chemical history for ozone time-series predictions-a method development study for deep learning models. EGU General Assembly Conference Abstracts, 19-30 April, pp. EGU21-12146.
17. Liu, H., Liu, J., Liu, Y., Ouyang, B., Xiang, S., Yi, K., Tao, S., 2020. Analysis of wintertime O3 variability using a random forest model and high-frequency observations in Zhangjiakou—an area with background pollution level of the North China Plain. Environmental Pollution, 262, 114191.
18. Liu, R., Ma, Z., Liu, Y., Shao, Y., Zhao, W., Bi, J., 2020. Spatiotemporal distributions of surface ozone levels in China from 2005 to 2017: A machine learning approach. Environment international, 142, 105823.
19. Liwicki, M.; Fernandez, S.; Bertolami, R.; Bunke, H.; Schmidhuber, J. (2009). "A Novel Connectionist System for Improved Unconstrained Handwriting Recognition". (IEEE Transactions on Pattern Analysis and Machine Intelligence. 31 (5): 855
20. Ma, R., Ban, J., Wang, Q., Zhang, Y., Yang, Y., He, M. Z., Li, S., Shi, W., Li, T., 2021. Random forest model based fine scale spatiotemporal O3 trends in the Beijing-Tianjin-Hebei region in China, 2010 to 2017. Environmental Pollution, 276, 116635.
21. Ma, Z., Liu, R., Bi, J., 2019. Spatiotemporal distributions of ground ozone levels in China from 2005 to 2016: a machine learning approach. AGU Fall Meeting Abstracts, 9-13 December 2019, San Francisco, USA, pp. A41J-2709.
22. Makarova, А., Evstaf'eva, E., Lapchenco, V., Varbanov, P. S., 2021. Modelling tropospheric ozone variations using artificial neural networks: A case study on the Black Sea coast (Russian Federation). Cleaner Engineering and Technology, 5, 100293.
23. Matasović, B., Pehnec, G., Bešlić, I., Davila, S., Babić, D., 2021. Assessment of ozone concentration data from the northern Zagreb area, Croatia, for the period from 2003 to 2016. Environmental Science and Pollution Research, 1-11.
24. Mehdipour, V., Memarianfard, M., 2019. Ground-level O3 sensitivity analysis using support vector machine with radial basis function. International Journal of Environmental Science and Technology, 16(6), 2745-2754.
25. Nghiem, T. D., Mac, D. H., Nguyen, A. D., Lê, N. C., 2021. An integrated approach for analyzing air quality monitoring data: a case study in Hanoi, Vietnam. Air Quality, Atmosphere & Health, 14(1), 7-18.
26. Paoli, C., Notton, G., Nivet, M. L., Padovani, M., Savelli, J. L. 2011. A neural network model forecasting for prediction of hourly ozone concentration in Corsica. 2011 10th International Conference on Environment and Electrical Engineering, 1-7 May 2011, Rome, Italy, pp. 1-4.
27. Sak, Hasim; Senior, Andrew; Beaufays, Francoise (2014). "Long Short-Term Memory recurrent neural network architectures for large scale acoustic modeling"
28. Sayeed, A., Choi, Y., Eslami, E., Jung, J., Lops, Y., Salman, A. K., Lee, J. B., Park, H. J., Choi, M. H. (2021). A novel CMAQ-CNN hybrid model to forecast hourly surface-ozone concentrations 14 days in advance. Scientific reports, 11(1), 1-8.
29. Sepp H., Jürgen S., 1997. Long short-term memory.
30. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R. (2014). Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15 (1), 1929–1958.
31. Şen, Z., 2018. Significance of Artificial Intelligence in Science and Technology. Journal of Intelligent Systems: Theory and Applications, 1(1), 1-4.
32. T. Chen, C. Guestrin, M. Assoc Comp, XGBoost: a scalable tree boosting system, 2016.
33. Tanaskuli, M., Ahmed, A. N., Zaini, N., Abdullah, S., Borhana, A. A., Mardhiah, N. A., 2020. Ozone prediction based on support vector machine. Indonesian Journal of Electrical Engineering and Computer Science, 17(3), 1461-1466.
34. Wang, H. W., Li, X. B., Wang, D., Zhao, J., & Peng, Z. R., 2020. Regional prediction of ground-level ozone using a hybrid sequence-to-sequence deep learning approach. Journal of Cleaner Production, 253, 119841.
35. Yang, X., Zhang, M., Zhang, B., 2021. A Generic Model to Estimate Ozone Concentration from Landsat 8 Satellite Data Based on Machine Learning Technique. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14, 7938-7947.
36. Yıldırım, A. E., Kadıoğlu, Ö. F., Kavak, H., Salman, K., Uçar, M. K., Uçar, Z., Bozkurt, M. R., 2021. Gender-Based Artificial Intelligence Based Detection of Basal Metabolic Rate by Electrocardiography Signal. Journal of Intelligent Systems: Theory and Applications, 4(2), 168-176.