Seul Şehri meteorolojik verileri kullanılarak makine öğrenmesi regresyon yöntemleri ile hava sıcaklığının değerlendirilmesi

Abstract

Havanın insan yaşamı ve faaliyetleri üzerinde önemli bir etkisi vardır. Hava sıcaklığındaki ani değişimler günlük yaşamı ve çeşitli endüstrileri olumsuz etkilendiğinden hava tahmini doğruluğunun önemini günden güne artırmaktadır. Mevcut hava tahmin yöntemleri iki ana gruba ayrılabilir: sayısal tabanlı ve makine öğrenim tabanlı yaklaşımlar. Sayısal tabanlı hava tahmin yöntemleri, hesaplama maliyetini önemli ölçüde artıran karmaşık matematiksel formüller kullanır. Buna karşın, makine öğrenim tabanlı yöntemler ise düşük işlem maliyetleri nedeniyle son yıllarda daha çok tercih edilir. Bu çalışmada, geleneksel makine öğrenmesi yöntemlerinin yanı sıra son yıllarda geliştirilen yükseltme tabanlı makine öğrenmesi algoritmaları ile birlikte 12 farklı regresyon yöntemi kullanılarak Güney Kore Seul için bir sonraki günün maksimum ve minimum hava sıcaklığı tahmin edilmektedir. Ayrıca, hiperparametrelerin ayarlanması, makine öğrenmesi algoritmalarının işlem süresini ve performansını etkilediğinden, 12 yöntemin tümü zaman ve hiperparametreler açısından kapsamlı bir şekilde çalışılmııştır. Yöntemlerin performanslarının karşılaştırılmasında literatürde sıklıkla tercih edilen kare korelasyon katsayısı (𝑅 2 ) kullanılmaktadır. Gözlemlenen sonuçlara göre, yükseltme tabanlı XGBoost ve LightGBM yöntemleri, hem istatistiksel test analizi hem de en yüksek 𝑅 2 puanı ile tüm yıllar için maksimum ve minimum hava sıcaklığını tahmin etmede en başarılı makine öğrenmesi algoritmalarıdır.

Keywords

• Makine öğrenmesi
• Günlük hava durumu tahmini
• Regresyon yöntemi
• Seul

Evaluation of air temperature with machine learning regression methods using Seoul City meteorological data

Abstract

Weather has a significant impact on human life and activities. As abrupt changes in air temperature negatively affect daily life and various industries, the importance of weather forecast accuracy is increasing day by day. Current weather forecasting methods can be divided into two main groups: numerical-based and machine learning-based approaches. Numerical-based weather forecasting methods use complex mathematical formulas that significantly increase the computational cost. On the other hand, machine learning-based methods have been preferred more in recent years due to their lower computational costs. In this study, the next day's maximum and minimum air temperature are estimated for Seoul, South Korea by using 12 different regression methods together with the boosting-based machine learning algorithms developed in recent years, as well as traditional machine learning methods. Furthermore, since tuning of hyperparameters affects the process time and performance of machine learning algorithms, all 12 methods have been extensively studied in terms of time and hyperparameters. The square correlation coefficient (𝑅 2 ), which is frequently adopted in the literature, is used to compare the performances of the methods. According to the observed results, the boosting-based XGBoost and LightGBM methods are the most successful machine learning algorithms in predicting the maximum and minimum air temperature for all years with both statistical test analysis and the highest 𝑅 2 score

Keywords

• Machine learning
• Daily weather forecasting
• Regression method
• Seoul

References

1. [1] Bushara NO, Abraham A. “Weather forecasting in Sudan using machine learning schemes”. Journal of Network and Innovative Computing, 2(1), 309-317, 2014.
2. [2] Holmstrom M, Liu D, Vo C. “Machine learning applied to weather forecasting”. Meteorological Applications, 10, 1-5, 2016.
3. [3] Saba T, Rehman A, AlGhamdi JS. "Weather forecasting based on hybrid neural model". Applied Water Science, 7(7), 3869-3874, 2017.
4. [4] Sharaff A, Roy SR. "Comparative analysis of temperature prediction using regression methods and back propagation neural network". IEEE 2018 2nd International Conference on Trends in Electronics and Informatics (ICOEI), Tirunelveli, India, 11-12 May 2018.
5. [5] dos Santos RS. "Estimating spatio-temporal air temperature in London (UK) using machine learning and earth observation satellite data". International Journal of Applied Earth Observation and Geoinformation, 2020. https://doi.org/10.1016/j.jag.2020.102066
6. [6] Ferreira LB, da Cunha FF. "New approach to estimate daily reference evapotranspiration based on hourly temperature and relative humidity using machine learning and deep learning". Agricultural Water Management, 2020. https://doi.org/10.1016/j.agwat.2020.106113
7. [7] Wolff S, O'Donncha F, Chen B. "Statistical and machine learning ensemble modelling to forecast sea surface temperature". Journal of Marine Systems, 2020. https://doi.org/10.1016/j.jmarsys.2020.103347
8. [8] Lee S, Lee YS, Son Y. "Forecasting daily temperatures with different time interval data using deep neural networks". Applied Sciences, 2020. https://doi.org/10.3390/app10051609

9. [9] Jakaria AHM., Hossain MM., Rahman MA. “Smart weather forecasting using machine learning: a case study in Tennessee”. arXiv, 2020. https://arxiv.org/pdf/2008.10789.pdf
10. [10] Akyüz AÖ, Kumaş K, Ayan M, Güngör A. “Antalya ili meteorolojik verileri yardımıyla hava sıcaklığının yapay sinir ağları metodu ile tahmini”. Gümüşhane Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 10(1), 146-154, 2020.
11. [11] Sevinç A, Kaya B. “Derin Öğrenme ve İstatistiksel Modelleme Yöntemiyle Sıcaklık Tahmini ve Karşılaştırılması”. Avrupa Bilim ve Teknoloji Dergisi, (28), 1222-1228, 2021.
12. [12] Cho D, Yoo C, Im J, Cha DH. "Comparative assessment of various machine learning‐based bias correction methods for numerical weather prediction model forecasts of extreme air temperatures in urban areas". Earth and Space Science, 2020. https://doi.org/10.1029/2019EA000740
13. [13] Duo D, Graff C. "UCI Machine Learning Repository". https://archive.ics.uci.edu/ml (15.01.2021).
14. [14] Hoerl AE, Kennard RW. "Ridge regression: biased estimation for nonorthogonal problems". Technometrics, 12(1), 55-67, 1970.
15. [15] Muthukrishnan R, Rohini R. "LASSO: A feature selection technique in predictive modeling for machine learning". IEEE 2016 International Conference on Advances in Computer Applications (ICACA), Coimbatore, India, 24-24 October 2016.
16. [16] Münch MM, Peeters CF, Van Der Vaart AW, Van De Wiel MA. "Adaptive group-regularized logistic elastic net regression". Biostatistics, 2018. https://doi.org/10.1093/biostatistics/kxz062
17. [17] González-Briones A, Hernández G, Pinto T, Vale Z, Corchado JM. "A review of the main machine learning methods for predicting residential energy consumption". IEEE 2019 16th International Conference on the European Energy Market (EEM), Ljubljana, Slovenia, 18-20 September 2019.
18. [18] Bottou L. "Large-scale machine learning with stochastic gradient descent". Proceedings of COMPSTAT'2010, 2010. https://doi.org/10.1007/978-3-7908-2604-3_16
19. [19] Imandoust SB, Bolandraftar M. "Application of k-nearest neighbor (knn) approach for predicting economic events: Theoretical background". International Journal of Engineering Research and Applications, 3(5), 605-610, 2013.
20. [20] Rhys HI. Machine Learning with R, the Tidyverse, and MLR. 1st ed. New York, USA, Manning, 2020.
21. [21] Yumus M, Apaydin M, Degirmenci A, Karal O. "Missing data imputation using machine learning based methods to improve HCC survival prediction". IEEE 2020 28th Signal Processing and Communications Applications Conference (SIU), Gaziantep, Turkey, 5-7 October 2020.
22. [22] Karasu S, Altan A. “Recognition model for solar radiation time series based on random forest with feature selection approach”. In 2019 11th International Conference on Electrical and Electronics Engineering (ELECO) IEEE, Bursa, Turkey, 28-30 November 2019.
23. [23] Vamsidhar E, Varma KVSRP, Rao PS, Satapati R. "Prediction of rainfall using backpropagation neural network model." International Journal on Computer Science and Engineering, 2(4), 1119-1121, 2010.
24. [24] Karal O. "Maximum likelihood optimal and robust Support Vector Regression with lncosh loss function". Neural Networks, 94, 1-12, 2017.
25. [25] Karal O. "EKG verilerinin destek vektör regresyon yöntemiyle sıkıştırılması". Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, 33(2), 743-756, 2018.
26. [26] Degirmenci A, Karal O. “Evaluation of kernel effects on svm classification in the success of wart treatment methods”. American Journal of Engineering Research, 7, 238-244, 2018.
27. [27] Freund Y, Schapire RE. “A decision-theoretic generalization of on-line learning and an application to boosting”. Journal of Computer and System Sciences, 55(1), 119-139, 1997.
28. [28] Ogunleye A, Wang QG. "XGBoost model for chronic kidney disease diagnosis". IEEE/ACM Transactions on Computational Biology and Bioinformatics, 17(6), 2131-2140, 2019.
29. [29] Ke G, Meng Q, Finley T, Wang T, Chen W, Ma W, Ye Q, Liu TY. “Lightgbm: A highly efficient gradient boosting decision tree”. Advances in Neural Information Processing Systems, 30, 3146-3154, 2017.
30. [30] Hacioğlu R. “Prediction of solar radiation based on machine learning methods”. The Journal of Cognitive Systems, 2(1), 16-20, 2017.
31. [31] Karal O. "Performance comparison of different kernel functions in SVM for different k value in k-fold crossvalidation". IEEE 2020 Innovations in Intelligent Systems and Applications Conference (ASYU), Istanbul, Turkey, 15-17 October 2020.
32. [32] García S, Fernández A, Luengo, J, Herrera, F. “Advanced nonparametric tests for multiple comparisons in the design of experiments in computational intelligence and data mining: Experimental analysis of power”. Information Sciences, 180(10), 2044-2064, 2010.