Comparative Evaluation of Hybrid SARIMA and Machine Learning Techniques Based on Time Varying and Decomposition of Precipitation Time Series

Document Type : Original Research

Author
L. Parviz
Faculty of Agriculture, Azarbaijan Shahid Madani University, Tabriz, Islamic Republic of Iran.
Abstract
Accurate precipitation forecasts are much attractive due to their complexity. This study aimed to use the hybrid Seasonal Autoregressive Integrated Moving Average (SARIMA) model and machine learning techniques such as Artificial Neural Networks (ANN) and Support Vector Machines (SVM) to improve precipitation forecasts. Time variation analysis and time series decomposition were the two concepts applied to construct the hybrid models. The performance of the two concepts was evaluated with monthly precipitation time series of two stations in northern Iran. Time variation analysis of time series was conducted with the clustering analysis, which increased the accuracy of forecasting with 20.99% decrease in the geometric mean error ratio for the two stations. SVM model decreased the forecasted error compared to ANN in the internal process of time variation analysis. Average of Mean Relative Error (MRE) were MRESVM= 0.72, MREANN= 0.89, and Mean Absolute Error (MAE) in the two stations were MAESVM= 18.02 and MAEANN= 23.88. Therefore, SVM outperformed the ANN model. Comparison of the two hybrid models indicated that more accurate results belonged to the concept of time series decomposition (the decrease in root mean square error from time variation to time series decomposition concepts was 13.35%). Extracting the pattern of data with SARIMA-based hybrid model with time series decomposition improved the precipitation forecasting. Configurations related to nonlinear components of time series with time steps of residual had good performance (the average of agreement index was 0.9). The results suggest that the hybrid model can be a valuable and effective tool for decision processes, and time series decomposition to linear and nonlinear components has a better performance.

Keywords

Subjects

1. Adamowski, J. and C. H. Karapataki. 2010. Comparison of Multivariate Regression and Artificial Neural Networks for Peak Urban Water-Demand Forecasting: Evaluation of Different ANN Learning Algorithms. J. Hydrol. Engin., 15: 729-743.
2. Bas, M., Ortiz, J., Ballesteros, L. and Martorell, S.2017. Evaluation of a multiple linear regression model and SARIMA model in forecasting 7Be air concentrations. Chemosphere. 177: 326-333.
3. Belaid, S. and A. Mellit. 2016. Prediction of daily and mean monthly global solar radiation using support vector machine in an arid climate. Energy Convers. Manag., 118: 105-118.
4. Box, G. E. P. and G. M. Jenkins. 1976. Times series analysis -forecasting and control. Prentice-Hall, Englewood Cliffs.
5. Box, G. E. P., Jenkins, G. M., Reinsel, G. C. and G. M. Ljung. 2015. Time series analysis: Forecasting and Control, John Wiley & Sons.
6. Chahkoutahi, F. and M. Khashei. 2017. Seasonal direct optimal hybrid model of computational intelligence and soft computing techniques for electricity load forecasting. Energy, 140: 988-1004.
7. Chen, K. Y. and C. H. Wang. 2007. A hybrid SARIMA and support vector machines in forecasting the production values of the machinery industry in Taiwan. Expert Syst. App., 32: 254-264.
8. Chen, X. and S. Zhu. 2013. Improved hybrid model based on support vector regression machine for monthly precipitation forecasting. J. Comput., 8(1): 232-238.
9. Cryer, J. D. and K. S. Chan. 2008. Time series analysis with application in R, 2nd edn. Springer, New York, p 491.
10. Diaz-Robles, L., Ortega, J. C., Fu, J. S., Reed, G. D., Chow, J. C., Watson, J. G. and J. A. Moncada-Herrera. 2008. A hybrid ARIMA and artificial neural network model to forecast particular matter in urban areas: the case of Temuco-Chile.
Atmos. Environ., 42: 8331-8340.
11. Du, J., Yayun, L., Yu, Y. and W. Yan. 2017. A prediction of precipitation data based on support vector machine and particles swarm optimization (PSO-SVM) algorithms. Algorithm, 10(75): 1-15.
12. Fang, T. and R. Lahdelma. 2016. Evaluation of a multiple linear regression model and SARIMA model in forecasting heat demand for district heating system. App. Energy, 179: 544-552.
13. Glisˇovic´, N., Milenkovic´, M., Bojovic´, N., Sˇvadlenka, L. and Z. Avramovic´. 2016. A hybrid model for forecasting the volume of passenger flows on Serbian railways. Operational Res., 16: 271-285.
14. Hamidi, O., Poorolajal, J., Sadeghifar, M., Abbasi, H., Maryanaji, Z., Faridi, H. R. and L. Tapak. 2014. A comperative study of support vector machines and artificial neural network for predicting precipitation in Iran. Theor. Appl. Climatol., 119: 723-731.
15. Jadhav, V., Chinnappa Reddy, B. V. and G. M. Gaddi. 2017. Application of ARIMA model for forecasting agricultural prices. J. Agr. Sci. Tech., 19: 981-992.
16. Jeong, K., Koo, C., and T. Hong. 2014. An estimation model for determination the annual energy cost budget in educational facilities using SARIMA (seasonal autoregressive integrated moving average) and ANN (artificial neural network). Energy, 71: 71-79.
17. Khandelwal, I., Adhikari, R., and G. H. Verma. 2105. Time series forecasting using Hybrid ARIMA and ANN models based on DWT decomposition. Procedia Comput. Sci., 48: 173-179.
18. Lee, N-UK, Shim, J-S., Ju, Y. W. and S-Ch. Park. 2018. Design and implementation of the SARIMA-SVM time series analysis algorithm for the improvement of atmospheric environment forecast accuracy. Soft Comput., 22: 4275–4281.
19. Liang, Y. U. 2009. Combining seasonal time series ARIMA method and neural networks with genetic algorithms for predicting the production value of the mechanical industry in Taiwan. Neural Comput. App., 18: 833-841.
20. Liu, Y, and Z. Ge. 2018. Weighted random forests for fault classification in industrial processes with hieratchical clustering model selection. J. Process Control, 64: 62-70.
21. Mo, L., Xie, L., Jiang, X., Teng, G., Xu, L. and J. Xiao. 2018. GMDH-based hybrid model for container throughput forecasting: selective combination forecasting in nonlinear subseries. App. Soft Comput., 62: 478-490.
22. Naguib, I. A. and H. W. Darwish. 2012. Support vector regression and artificial neural network models for stability indicating analysis of mebeverine hydrochloride and sulpiride mixtures in pharmaceutical preparation: A comparative study. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. Spectrochimica Acta Part A., 86: 515-526
23. Narasimha Murthy, K. V., Saravana, R. and K. Vijaya Kumar. 2018. Modeling and forecasting rainfall patterns of southwest monsoons in North–East India as a SARIMA process. Meteorol. Atmos. Phys., 130: 99-106.
24. Niedbala, G. and R. J. Kozlowski. 2019. Application of Artificial Neural Networks for Multi-Criteria Yield Prediction of Winter Wheat. J. Agr. Sci. Tech., 21: 51-6.
25. Papacharalampous, G., Tyralis, H. and D. Koutsoyiannis. 2018a. One-step ahead forecasting of geophysical processes within a purely statistical framework. Geosci. Lett., 5(12).
26. Papacharalampous, G., Tyralis, H. and D. Koutsoyiannis. 2018b. Predictability of monthly temperature and precipitation using automatic time series forecasting methods. Acta Geophysic., 66(4): 807–831.
27. Papacharalampous, G., Tyralis, H. and D. Koutsoyiannis. 2018c. Univariate time series forecasting of temperature and precipitation with a focus on machine learning algorithms: A multiple-case study from Greece. Water Resourc. Manage., 32(15): 5207–5239.
28. Papacharalampous, G. and H.Tyralis. 2018. Evaluation of random forests and Prophet for daily streamflow forecasting. Adv. Geosci., 45: 201–208.
29. Papacharalampous, G., Tyralis, H. and D. Koutsoyiannis. 2019. Comparison of stochastic and machine learning methods for multi-step ahead forecasting of hydrological processes. Stochastic Environ. Res. Risk Assess., 1-34.
30. Rathod, S and G. C. Mishra. 2018. Statistical Models for Forecasting Mango and Banana Yield of Karnataka, India. J. Agr. Sci. Tech., 20: 803-816
31. Ruiz-Aguilar, J. J., Turias, I. J., Jimenez-Come, M. J. and M. Mar Cerban, 2014. Hybrid Approaches of support vector regression and SARIMA models to forecast the inspections volume. Int. Conf. Hybrid Artificial Intelligence Syst., 502-514.
32. Selvanayaki, K. S. and R. Somasundaram. 2015. An improved approach for detection and classification of vehicles in video using support vector machines. ARPN J. Engin. App. Sci., 10(10): 4690-4700.
33. Taghadomi-Saberi, S. and S. J. Razavi.2019. Evaluating Potential of Artificial Neural Network and Neuro-Fuzzy Techniques for Global Solar Radiation Prediction in Isfahan, Iran. J. Agr. Sci. Tech., 21(2): 295-307.
34. Tealab, A., Hefny, H. and A. Badr. 2017. Forecasting of nonlinear time series using artificial neural network. Future Comput. Informatics J., 1: 9.
35. Theil, h. 1961. Economic forecasts and policity. North –Holland Pub. Co.
36. Theil, h. 1966. Applied economic forecasting. North- Holland Pub. Co.
37. Tyralis, H. and G. Papacharalampous. 2018. Large-scale assessment of Prophet for multi-step ahead forecasting of monthly streamflow. Adv. Geosci., 45: 147–153.
38. Tyralis, H. and G. Papacharalampous. 2017. Variable selection in time series forecasting using random forests. Algorithms, 10(4): 114.
39. Vapnik, V. N. 1995. The nature of statistical learning theory. New York: Springer.
40. Wang, H. R., Wang, C., Lin, X. and J. Kang, 2014. An improved ARIMA model for precipitation simulations. Nonlin. Processes Geophys., 21: 1159-1168.
41. Wang, H. R., Ye, L. T. and C. M. Liu. 2007. Problems in wavelet analysis of hydrologic series and some suggestion on improvement. Prog. Nat. Sci., 17: 80-86.
42. Weng, C., Huang, T. and R. Han. 2016. Disease prediction with different types of neural network classifiers. Telematic. Inform., 33:277–292.
43. Willmott, C. J. 1981. On the validation of models. Phys. geograph., 2: 184-194.
44. Yolcu, U., Egrioglu, E. and C. H. Aladag. 2013. A new linear and nonlinear artificial neural network model for time series forecasting. Decision Support Syst., 54: 1340-1347.
45. Zaynoddin, M., Bonakdari, H., Azari, A., Ebtehaj, I., Gharabaghi, B. and H. Riahi Madavar. 2018. Novel hybrid linear stochastic with non-linear extreme learning machine methods for forecasting monthly rainfall a tropical climate. J. Environ. Manage., 222: 190-206.
46. Zhang, G. 2003. Time series forecasting using a hybrid ARIMA and neural network model. Neurocomput., 50: 159-175.
47. Zhang, J., Wei, Y. M., Li, D., Tan, Z. and J. Zhou. 2018. Short term electricity load forecasting using a hybrid model. Energy, 158: 774-781.
48. Zhu, S., Lian, X., Liu, H., Hu, J., Wang, Y. and J. Che. 2017. Daily air quality index forecasting with hybrid models: A case in China. Environ. Pollut., 231: 1232-1244.
Journal of Agricultural Science and Technology
Volume 22, Issue 2
March and April 2020
Pages 563-578