Forecasting accuracy and change point detection

Gregory Gurevich, Yossi Hadad, Baruch Keren

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The accuracy of time series forecasting often decreases because of the existence of change points in the data. This paper presents a novel method for time series forecasting that taking into account the possibility of a change point in past data. The proposed method can be applied to situations where the considered time series consists of independent or weakly dependent observations. Change point analysis prevents the omission of relevant data as well as the forecasting that may be based on irrelevant data. The study demonstrates that change point techniques may increase the accuracy of forecasts.

Original languageEnglish
Title of host publicationProceedings of the 14th International Symposium on Operational Research, SOR 2017
EditorsSamo Drobne, Janez Zerovnik, Lidija Zadnik Stirn, Mirjana Kljajic Borstar
PublisherSLOVENIAN SOCIETY INFORMATIKA
Pages314-319
Number of pages6
ISBN (Electronic)9789616165501
StatePublished - 1 Jan 2017
Externally publishedYes
Event14th International Symposium on Operational Research, SOR 2017 - Bled, Slovenia
Duration: 27 Sep 201729 Sep 2017

Publication series

NameProceedings of the 14th International Symposium on Operational Research, SOR 2017
Volume2017-September

Conference

Conference14th International Symposium on Operational Research, SOR 2017
Country/TerritorySlovenia
CityBled
Period27/09/1729/09/17

Keywords

  • Business forecasting
  • Change point
  • Error indexes
  • Homogeneous series

ASJC Scopus subject areas

  • Management Science and Operations Research
  • Strategy and Management
  • Modeling and Simulation
  • Management of Technology and Innovation
  • Numerical Analysis
  • Statistics, Probability and Uncertainty
  • Computational Theory and Mathematics

Fingerprint

Dive into the research topics of 'Forecasting accuracy and change point detection'. Together they form a unique fingerprint.

Cite this