Analysis and Forecasting of Parameters for p-Order Autoregressive Process

Authors

  • A. A. Sherstneva Siberian State University of Telecommunications and Information Sciences, Novosibirsk

DOI:

https://doi.org/10.22213/2413-1172-2020-4-77-84

Keywords:

regression, analysis, least squares estimation, Yule-Walker equations, Levinson-Durbin algorithm

Abstract

Several assumptions are made while solving the problems of calculating the performance and reliability metrics of info-communication systems using classical methods – for instance, the exponential distribution of source variables does not always correspond to real data. Moreover, source variables are random variables collected and processed by the monitoring system. To obtain the most accurate results of the indicator calculation, it is necessary to conduct a large number of measurements of random variables. In this sense, the well-known formulas of teletraffic theory have some inaccuracy. One of the effective ways to solve this problem is the use of regression analysis. The paper purposes for estimating the metrics in info-communication systems for data trend forecasting.

The paper aims to identify the autoregressive process for parameters forecasting for info-communication systems. The use of a classical regression analysis method, such as least squares estimation, has several limitations. The paper proposes an alternative approach to solving the indicated problem through the filtration theory.

Along with the classical least squares estimation, an alternative opportunity is to solve the empirical Yule-Walker equations. The paper presents a solution technique using the Levinson-Durbin algorithm. In addition to theoretical calculations, a program has been developed to automate the computation process.

Depending on the number of observations, a criterion is selected, its minimization leads to the model’s real order. The estimation is carried out using the mathematical modeling program Matlab. In this regard, the paper considers the possibility of choosing a criterion from a theoretical point of view and its practical implementation. The conditions of use are also given; the most effective method for criterion choosing is shown depending on the number of observations.

References

Шерстнева А. А., Шерстнева О. Г. Анализ сети связи с учетом показателей надежности // Вестник Рязанского гос. радиотехнического ун-та. 2020. № 73 (2). С. 52–58.

Домбровский В. В., Объедко Т. Ю. Оптимальные стратегии прогнозирующего управления системами со случайными параметрами, описываемыми многомерной регрессионной моделью с марковским переключением режимов // Вестник Томского гос. ун-та. 2019. № 48. С. 4–12.

Домбровский В. В., Пашинская Т. Ю. Прогнозирующее управление системами с марковскими скачками и авторегрессионным мультипликативным шумом с марковским переключением режимов // Вестник Томского гос. ун-та. 2018. № 44. С. 4–9.

Афанасьев А. А., Власов Р. С. Особенности выделения сегментов анализа речевого сигнала при его обработке // Перспективные технологии в средствах передачи информации : cборник тр. конф. / Владимирский гос. ун-т, 2019. С. 156–160.

Моттль В. В., Красоткина О. В., Ежова Е. О. Непрерывное обобщение информационного критерия Акаике для оценивания нестационарной регрессионной модели временного ряда с неизвестной степенью изменчивости коэффициентов // Математические методы распознания образов. 2009. № 1. С. 52–55.

Вилков А. П., Родионова Т. Е. Использование систем одновременных уравнений для получения моделей описания технических объектов // Современные проблемы проектирования, производства и эксплуатации радиотехнических систем. 2016. № 10. С. 175–177.

Св. о регистрации электронного ресурса № 17760 / О. Г. Шерстнева, А. А. Шерстнева. Программа имитации функционирования телекоммуникационной сети с учетом реальных показателей надежности. 29.12.2011 г.

Wickham H. Elegant graphics for data analysis. 2nd ed. Springer, 2016, 213 p.

Athanasopoulos G., Hyndman R.J., Kourentzes N., Petropoulos F. Forecasting with temporal hierarchies. European J. of Operational Research, 2017, no. 262, pp. 60-74.

Bergmeir C., Hyndman R.J., Benítez J.M. Bagging exponential smoothing methods using STL decomposition and Box-Cox transformation. International J. of Forecasting, 2016, no. 32, pp. 303-312.

Sherstneva A., Sherstneva O. Analysis statistical information for data trend forecasting. International Ural conference on Electrical power Engineering. IEEE, 2020, pp. 153-158.

Кантарович Г. Г. Экономическая статистика. Эконометрика. М. : ГУ-ВШЭ, 2000.

Judge G.G., Griffits W.E., Hill R.C. The theory and practice of econometrics. NY, John Willey and Sons, 1985.

Bergmeir C., Hyndman R.J., Koo B. A note on the validity of cross-validation for evaluating autoregressive time series prediction. Computational Statistics and Data Analysis, 2018, no. 120, pp.70-83.

Wickramasuriya S.L., Athanasopoulos G. Optimal forecast reconciliation for hierarchical and grouped time series through trace minimization. J. American Statistical Association, 2019, no. 114, pp. 804-819.

Harrell F.E. Regression modeling strategies: With applications to linear models, logistic and ordinal regression, and survival analysis. 2nd ed. NY, Springer, 2015, 568 p.

Madsen K., Nielsen H.B., Tingleff O. Methods for Non-linear Least Squares Problem Cobenhavn. Technical University of Denmark, 2004, 30 p.

Published

30.12.2020

How to Cite

Sherstneva А. А. (2020). Analysis and Forecasting of Parameters for p-Order Autoregressive Process. Vestnik IzhGTU Imeni M.T. Kalashnikova, 23(4), 77–84. https://doi.org/10.22213/2413-1172-2020-4-77-84

Issue

Section

Articles