FORECASTING THE STATE OF TELECOMMUNICATION NETWORKS USING QUANTILE AND LOGISTIC REGRESSION METHODS
DOI:
https://doi.org/10.17721/2519-481X/2023/80-10Keywords:
information technology, telecommunication system, intellectual decision-making system, machine learning, quantile regressionAbstract
In today's modern world, the ubiquity of information technology has intertwined telecommunications systems with every facet of human life. It's challenging to fathom a world where you're disconnected from the "World Wide Web" or unable to exchange data instantly via the intricate web of modern mobile devices. The vitality of staying connected online cannot be overstated, and ensuring the smooth functioning of telecommunications systems is paramount. This paper delves into the pivotal task of predicting and managing the performance of these networks, employing quantile and logical regression techniques. Our study leverages real-world data from telecommunication network operations to construct a predictive model capable of anticipating network conditions in advance. This predictive capability serves as the linchpin for intelligent decision-making systems, facilitating real-time network management. By implementing machine learning methods, specifically quantile regression, we achieve a sophisticated understanding of how various factors influence network performance. Our research doesn't just stop at forecasting; it extends to the realms of intellectual decision-making systems, where the insights gained from regression analysis play a pivotal role. These intelligent systems are equipped to make data-driven decisions on network resource allocation, maintenance schedules, and preemptive problem resolution. In essence, they act as the custodians of network stability, ensuring that telecommunication systems remain robust and responsive to the ever-evolving demands of modern society. This article sheds light on the indispensable role of regression methods in proactively managing the state of telecommunications networks. By harnessing the power of machine learning and data-driven insights, we pave the way for a future where network disruptions are minimized, and the seamless connectivity we've come to rely on remains a constant presence in our lives.
References
1. Khlaponin Y.I. Modern problems of creating complex real-time information and control systems /Khlaponin Yu.I., Nedaybida Yu.P., Kotova Yu.V. // Information protection. – 2012. – No. 4 (57). – P. 50-55.
2. Enyukov I.S. Statistical analysis and monitoring of scientific and educational Internet networks /I.S. Enyukov, I.V. Retinskaya; under. ed. A. N. Tikhonova. – M.: Finance and Statistics, 2004. – 320 p.
3. Khlaponin Yu.I. Modern problems of creating complex real-time information and control systems/ Khlaponin Yu.I., Nedaybida Yu.P., Kotova Yu.V. // Information protection. – 2012. – No. 4 (57). – P. 50-55.
4. Ayvazyan S.A. Applied statistics: classification and dimensionality reduction / Ayvazyan S.A.,Bukhstaber V.M., Enyukov I.S., Meshalkin L.D. Applied statistics: classification and dimensionality reduction. – M.: Finance and Statistics, 1989.
5. Andrews D.F. A Robust Method for Multiple Linear Regression // Technometrics. 1974. Vol. 16.P.523–531.
6. Hogg R.V. Adaptive Robust Procedures: A Partial Review and Some Suggestions for FutureApplications and Theory // Journal of the American Statistical Association. 1972. Vol. 43. P. 1041–1067.
7. Khlaponin Y.I. Establishment of neural measures in a statistical system of analysis andmonitoring of telecommunication measures / Khlaponin Y.I., Zhirov G.B., Nikitchin O.M. - K.: Technological audit and production reserves, 2016. - P.35.
8. Koenker R., Bassett G., Jr. Regression Quantiles // Econometrica. 1978. Vol. 46. No.1.P.33–50.
9. Koenker R., Hallock K.F. Quantile Regression // Journal of Economic Perspectives. 2001.Vol. 15. No.4. P.143–156.
10.Huber P.J., Robust Statistics. 1981. New York: John Wiley and sons.
11. Karst O.J. Linear Curve Fitting Using Least Deviations // Journal of the AmericanStatistical Association. 1958. Vol. 53. No. 281. P. 118–132.
12. Wagner H.M. Linear Programming Techniques Regression Analysis // Journal of theAmerican Statistical Association. 1959. Vol. 54. No. 285. P.206–212.
13. Introduction to the theory of order statistics / Ed. A.E. Sarkhan and B.G. Greenberg. Per. fromEnglish Ed. AND I. Boyarsky. M.: Statistics, 1970.
14. Kenneth Q. Z., Stephen L. P. Direct Use of Regression Quantiles to Construct Confidence Setsin Linear Models // The Annals of Statistics. 1996. Vol. 24. No. 1. P. 287–306.
15. Serhey Lienkov, Genadiy Zhyrov, Ihor Pampukha, Ivan Chetverikov . Block EncryptionAlgorithm for Digital Information Using Open Keys for Selfgeneration of Closed Random Private Keys // 2019 IEEE International Conference on Advanced Trends in Information Theory (ATIT), 18-20 Dec. 2019 Kyiv, Ukraine. Electronic ISBN: 978-1-7281-6144-0, Print on Demand(PoD) ISBN: 978-1-7281-6145-7, INSPEC Accession Number: 19452664 DOI: 10.1109/ATIT49449.2019.9030509.
