Comparison of neural networks and regression time series in predicting development of EU export to PRC
Keywords:
neural networks, regression, time series, predicting development, exportAbstract
Purpose of the article This article focuses on the development of EU-China exports and trade in general, and also the advanatges and disadvantages of regression analysis and neural networks in prediction.
Methodology/methods The data for the analysis are available at the World Bank websites, etc. For the purpose of the analysis, the data on the EU export to the People´s Republic of China (hereinafter referred to as “PRC”) is used. The time period for which the data are available is a monthly export for the period starting from January 2000 and ending in July 2018. In total, it is 223 input data. The unit is euro.
Scientific aim The objective of the contribution is to compare the accuracy of equalized time series by means of regression analysis and neural networks on the example of the EU export to the PRC.
Findings It can be stated that due to the high simplification of the reality it is not possible to predict extraordinary situations and their impact on the EU export to the PRC (at least not in the long term). Prediction in the order of days would be ideal; however, it is not possible to obtain data for such a short prediction.
Conclusions The EU export to the PRC can be identified based on statistical, causal and intuitive methods. These, however, provided only a possible framework of the monitored variable development. It is important to work with the information on a possible future develop-ment of political, economic or legal environment. Optically, in terms of the linear regression, the most suitable one appear to be the LOWESS curve. The second best possible is the curve obtained by the negative exponential smoothing least squares method and distance-weighting least squares method. In terms of neural networks, all retained neural structures appear to be applicable in practice, as there is no significant difference between them.
References
Canofari, P., & Del Ponte, A. (2018). Chinese and European Financial Systems: Instability Drivers and Contagi-on Channels. International Advances in Economic Research, 24(4), 311-324.
Caprotti, F. (2015). Golden sun, green economy: market security and the US/EU-China ‘solar trade war’. Asian Geographer, 32(2), 99-115.
Cardoso, A., & Duarte, A. (2017). The impact of the Chinese exchange policy on foreign trade with the Europe-an Union. Brazilian Journal of Political Economy, 37(4), 870-893.
Curran, L., Lv, P., & Spigarelli, F. (2017). More heat than light? Renewable energy policy and EU–China solar energy relations. International Journal of Ambient Energy, 38(5), 471-480.
Gabriel, J., & Schmelcher, S. (2018). Three scenarios for EU-China relations 2025. Futures, 97, 26-34.
Goetz, L., & Grethe, H. (2010). The entry price system for fresh fruit and vegetable exports from China to the EU-Breaking a fly on the wheel? China Economic Review, 21(3), 377-393.
Karkanis, D. (2018). EU-China Trade: Geography and Institutions from 2001 to 2015. Journal of Economic Integration, 33(1), 1158-1175.
La Delfa, N. J., & Potvin, J. R. (2016). Predicting manual arm strength: A direct comparison between artificial neural network and multiple regression approaches. Journal of biomechanics, 49(4), 602-605.
Le Caillec, J. M., & Garello, R. (2003). Time series nonlinearity modeling: A Giannakis formula type approach. Signal Processing, 83(8), 1759-1788.
Mau, K. (2017). US Policy Spillover (?)–China’s Accession to the WTO and Rising Exports to the EU. European Economic Review, 98, 169-188.
Merdun, H., et al. (2006). Comparison of artificial neural network and regression pedotransfer functions for pre-diction of soil water retention and saturated hydraulic conductivity. Soil and Tillage Research, 90(1-2), 108-116.
Spigarelli, F., Arteconi, A., & Curran, L. (2017). Europe–China cooperation and competition in the green industries. International Journal of Ambient Energy, 38(5), 443-443.
Tao, F., et al. (2018). Driving forces of energy embodied in China-EU manufacturing trade from 1995 to 2011. Resources, Conservation and Recycling, 136, 324-334.
Tu, J. V. (1996). Advantages and disadvantages of using artificial neural networks versus logistic regression for predicting medical outcomes. Journal of clinical epidemiology, 49(11), 1225-1231.
Voituriez, T., & Wang, X. (2015). Real challenges behind the EU–China PV trade dispute settlement. Climate Policy, 15(5), 670-677.
Vrbka, J., & Rowland, Z. (2017). Stock price development forecasting using neural networks. SHS Web of Conferences, 39.
Zekić-Sušac, M., et al. (2016). Predicting company growth using logistic regression and neural networks. Cro-atian Operational Research Review, 7(2), 229-248.
Downloads
Published
Issue
Section
License
One of our top priorities and responsibilities is to prevent any illegal or unethical practice. Plagiarism is not tolerated in any form. Through the submission of a manuscript, the author confirms that it is their own original research. The ethical rules are binding for all parties who participate in the conference and are published in the Conference Proceedings. All parties to the editorial process respect the Ethical Rules of Conference.
