Seasonal decomposition and forecasting of telecommunication data: A comparative case study

A Recurrent Neural Network–based Forecasting System for Telecommunications Call Volume

“…These methods can be divided into three major categories: The first method employed is the Naive Forecast 1 [1], which uses the most recent observation as a forecast for the next time interval. A method that takes into account the seasonal factors is the Linear Extrapolation with Seasonal Adjustment (LESA, [2]): After seasonality has been removed from the original data, linear extrapolation is applied in order to forecast the future values of the series by employing the trend-cycle component. Finally, the projected trend-cycle component is adjusted, making use of the identified seasonal factors.…”

“…It presumes weak stationarity, equally spaced intervals or observations, and at least 30 to 50 observations. The well-established methods mentioned above have been studied in [2]. Linear models are also suggested by the ITU Recommendation E.507 for trend forecasting in telecommunications data [7].…”

Forecasting of telecommunications time-series via an Orthogonal Least Squares-based fuzzy model

“…Calls to national destinations comprise almost half the volume of the total outgoing calls from the campus. In the past, the forecasting ability of well established statistical methods on the University's call traffic has been studied [1]. Linear models have also been suggested for forecasting trends in telecommunications data by the ITU Recommendation E.507 [2].…”

Artificial intelligence diagnosis algorithm for expanding a precision expert forecasting system

“…Management judgment is normally required to adjust statistical forecasts in response to special events. Thus, forecasting methodology is most relevant and applicable in the field of management; moreover, a key objective of management is demand forecasting (Hilas, Goudos, & Sahalos, 2006;Lee, Cho, Lee, & Lee, 2006;Lee, Lee, & Kim, 2008). Short-term historic and stochastic volatility data are important for accurate forecasting.…”