2017 | OriginalPaper | Buchkapitel
Online Workload Forecasting
verfasst von : Nikolas Herbst, Ayman Amin, Artur Andrzejak, Lars Grunske, Samuel Kounev, Ole J. Mengshoel, Priya Sundararajan
Erschienen in: Self-Aware Computing Systems
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
This chapter gives a summary of the state-of-the-art approaches from different research fields that can be applied to continuously forecast future developments of time series data streams. More specifically, the input time series data contains continuously monitored metrics that quantify the amount of incoming workload units to a self-aware system. It is the goal of this chapter to identify and present approaches for online workload forecasting that are required for a self-aware system to act proactively—in terms of problem prevention and optimization—inferred from likely changes in their usage. The research fields covered are machine learning and time series analysis. We describe explicit limitations and advantages for each forecasting method.