Document Type : Research Paper

Authors

1 Assistant Professor, Department of Mathematics, Faculty of Basic Sciences, Ayatollah Boroujerdi University, Boroujerd, Iran

2 Assistant Professor, Department of Economics, Faculty of Human Sciences, Ayatollah Boroujerdi University, Boroujerd, Iran

Abstract

 One of the most challenging issues in forecasting economic variables is the lack of sufficient data or  the missing data in time series. In this paper, the time series of the GDP growth rate from 1980 to 2019 for 18 Middle East and North African countries is modeled via a generalized network autoregressive model. Of the total observations, 13.42% were missed. In the proposed model, a random network is applied to the data for which nodes represent countries or related time series. Then, an autoregressive model of each node was constructed based on all the data of its multi-stage neighboring nodes. Some parameters of the model may depend on the node (local model) or can be considered the same for all network nodes (global model). The missing data are modeled by changing the weights of the network edges. Finally, the time series was predicted based on the constructed model. Because the network structure affects the model and ultimately the forecast, and on the other hand it is difficult to examine all possible networks, ten thousand un-directional random networks and 16 models including 8 local models and 8 global models on each network are considered. Out of 160,000 models, the network and the model with the least prediction error are selected as the best network and model that are used for the main prediction. The lowest in-sample predictive error was obtained at a local network, which has 64 edges and the number of corresponding model parameters is 4. Finally, the model is compared with the classical models such as AR and VAR. The results indicate the superiority of the proposed method in significantly reducing the prediction error over the AR and VAR models.

Keywords

اسدی، غلامحسین و نقدی، سجاد. (1397). طراحی و تبیین الگوی پیش‌بینی رشد اقتصادی با رویکرد حسابداری. مجله دانش حسابداری، (3)9، 39-63.
برکچیان، سید مهدی، سمائی، کیان و نجفی زیارانی، فاطمه. (1400). ساخت نشانگر پیشرو ترکیبی برای پیش‌بینی تولید ناخالص داخلی ایران. مطالعات اقتصادی کاربردی ایران، (40)10، 39-71.
شایگانی، بیتا، سلامی، امیربهداد و خوچیانی، رامین. (1393). مدل پیشنهادی برای پیش بینی تولید ناخالص داخلی کاربرد مدل‌هایARIMA شبکه‌های عصبی و تبدیل موجک. دانش مالی تحلیل اوراق بهادار، (24)7، 147-162.
فاضل، مهدی، توکلی، اکبر و رجبی، مصطفی. (1392). مقایسه عملکرد الگوی ARIMA و MS-AR در پیش‌بینی ادوار تجاری ایران. مدلسازی اقتصادی، (22)7، 63-81.
مشکانی، محمدرضا و فخاری، علی. (1384). مقایسه کارایی مدل‌های کلاسیک وپویای بیزی در کاربردی از مدل‌های و پویای سری زمانی بیزی. تحقیقات اقتصادی، (4)40، 321-354.
Assadi (Ph.D), G., & Naghdi, S. (2018). Designing and Formulating the Forecasting Model of Economic Growth by Accounting Approach. Journal of Accounting Knowledge, 9(3), 39-63. doi: 10.22103/jak.2018.11095.2524 [In Persian]
Barakchian, S. M., Samaei, K., & Najafi Ziarani, F. (2022). Constructing a Composite Leading Indicator for Forecasting Non-Oil GDP. Journal of Applied Economics Studies in Iran, 10(40), 39-71. doi: 10.22084/aes.2021.20365.2957 [In Persian]
Bashir F. & Wei H. L. (2016). Handling Missing Data in Multivariate Time Series Using a Vector Autoregressive Model Based Imputation (VAR-IM) Algorithm. In Proceedings of the 24th Mediterranean Conference on Control and Automation (MED), 611–616. IEEE, Athens, Greece.
Erdös, P., & Rényi, A. (1959). On Random Graphs. I., Publicationes Mathematicae 6, 290–297.
Fazel, M., Tavakoli, A., & Rajabi, M. (2013). Comparing the Performance of ARIMA and MS-AR Models to Forecast Business Cycles in Iran. Economical Modeling, 7(22), 63-81. [In Persian]
Hanneke, S., Fu, W., Xing, E. P., Discrete temporal models of social networks. Electron. J. Statist. 4 585 - 605, 2010.
Knight, M., Leeming, K., Nason, G., & Nunes, M. (2020). Generalized network autoregressive processes and the GNAR package, Journal of Statistical Software, 96(5), 1-36.
Knight, M.I., Nunes, M.A., & Nason, G.P. (2016). Modelling, Detrending and Decorrelation of Network Time Series. arXiv: Methodology.
Krampe, J. (2019). Time series modeling on dynamic networks, Electronic Journal of Statistics, Electronic Journal of Statistics, 13(2), 4945-4976.
Krivitsky, P. N. and Handcock, M. S. (2014). A separable model for dynamic networks. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 76(1), 29-46.
Lütkepohl, H. (2005). New Introduction to Multiple Time Series Analysis. Berlin: Springer-Verlag.
Meshkani, M. & Fakhari, A. (2006). Comparing the performance of classical and dynamic Bayesian models in the application of Bayesian time series and dynamic models. Journal of Economic Research (Tahghighat- E- Eghtesadi), 40(4), 321-354. [In Persian]
 OECD (2021). Quarterly GDP.
Schwarz, G. (1978). Estimating the Dimension of a Model. The Annals of Statistics, 6(2), 461–464.
Shaygani, B., Salami, A., & Khochiani, R. (2014). The Proposed Model For Prediction Of GDP Using With ARIMA, Neural Networks And Wavelet Transform. Financial Knowledge of Securities Analysis, 7(24), 147-162. [In Persian]
Xu, K. (2015). Stochastic block transition models for dynamic networks. Artificial Intelligence and Statistics, 1079-1087.
Zellner, A. (1971). An Introduction to Bayesian Inference in Econometrics. Wiley, New York.
Zhu X., Pan R., Li G., Liu Y., Wang H. (2017). Network Vector Autoregression. The Annals of Statistics, 45, 1096–1123.