Generalized Cost-Function-Based Forecasting for Periodically Measured Nonstationary Traffic

Abstract-- In this paper, we address the issue of forecasting for periodically measured nonstationary traffic based on statistical time series modeling. Often with time series based applications, minimum mean square error (MMSE) based forecasting is sought that minimizes the square of the positive as well as the negative deviations of the forecast from the unknown true value. However, such a forecasting function is not directly applicable for applications such as predictive bandwidth provisioning in which the negative deviations (under-forecast) have more impact on the system performance than the positive deviation (overforecast). For instance, an under-forecast may potentially result in insufficient allocation of bandwidth leading to short term data loss. To facilitate a differential treatment between the under and the over-forecasts, we introduce a generalized forecast cost function that is defined by allowing different penalty associated with the under and the over-forecasts. We inv...

• Type: Journal
• Year: 2008
• Where: TSMC
• Authors: Balaji Krithikaivasan, Yong Zeng, Deep Medhi
• Keywords: Negative Deviations, Penalty Functions, Time Series, TSMC 2008