 On Impact of Data Models on Predictability Assessment of Time SeriesSergey Frenkel ()
A chapter in Data Analysis and Optimization, 2023, pp 103-134 from Springer Abstract: Abstract The main methods for predicting traffic in telecommunications networks are the methods and techniques of Machine Learning (ML). Within the framework of the ML approach, the network traffic predictor is considered as a tool that uses one way or another accumulated statistics over time to draw conclusions about the future behavior of network traffic. However, as the analysis of the literature shows, many modern MO tools do not work efficiently enough due to the pronounced non-linearity of traffic changes and non-stationarity. This paper is an attempt at some ordering and categorization in a huge stream of publications on modern methods, techniques and models for predicting data of various nature. For this, the concept of a conceptual model (CM) of algorithms for predicting the state of systems in the subject area is introduced. This concept is used widely used in modern Big Date and Software Engineering, is adapted here to represent specific probabilistic models of random sequences and processes of various nature (symbolic, binary, real). Under such a conceptual model, the paper refers to a formalized description of the relationships between elements, objects (both real and model) and goals of mathematical models and prediction algorithms, as well as the relationships between them, expressed in probabilistic terms. This construction allows representing classes of models according to the types of predicted data, used probability measures, loss functions when making decisions about the acceptability of the received forecast. Among the tasks of forecasting, the task of predicting signs of increments (direction of change) of the time series process is singled out separately. The paper proposes to use some results of the theory of random processes for a quick assessment of the predictability of signs of increments with acceptable ac-curacy. The proposed procedure is a simple heuristic rule for predicting the increment of two neighboring values of a random sequence. The connection with this approach for time series with known approaches for predicting binary sequences is shown. Date: 2023 There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it. Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:spr:spochp:978-3-031-31654-8_7 Ordering information: This item can be ordered from DOI: 10.1007/978-3-031-31654-8_7 Access Statistics for this chapter More chapters in Springer Optimization and Its Applications from Springer |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.