Intelligent Time Series Forecasting Through Neighbourhood Search Heuristics
Automated forecasting is essential to business operations that handle scores of univariate time series. Practitioners have to deal with thousands of time series with a periodicity ranging from seconds to monthly. The sheer velocity and volume of time series make it challenging for human labour to manually identify the order of the time series to forecast the results. An automated forecasting algorithm or framework is essential to complete the task. The approach must be robust in the identification of the order of the time series, and readily applicable to scores of time series without manual intervention. The most modern automated forecasting algorithms are derived from exponential smoothing or ARIMA models. In this paper, the authors proposed a new heuristics approach to identify the initial starting point for a neighbourhood search to obtain the most appropriate model. The results of this method are used to compare against the methods proposed in the literature.
KeywordsData mining Time series Forecasting Heuristics ARIMA
We wish to thank Michele Hibon and Makridakis Spyros for providing M3 data; Hyndman [14, 15] for his R package, Forecast which provides us with the Auto.Arima function, and Mcomp, which allowed us to implement M3 time series data easily in R; and the anonymous referees for insightful comments and suggestions.
- 6.Keogh, E., Lin, J., Fu, A.: HOT SAX: efficiently finding the most unusual time series subsequence. In: The Fifth IEEE International Conference on Data Mining (2005)Google Scholar
- 8.Gomez, V.: Automatic model identification in the presence of missing observations and outliers, Technical report, Ministerio de Economía y Hacienda, Dirección General de Análisis y Programación Presupuestaria, working paper D-98009 (1998)Google Scholar
- 9.Gomez, V., Maravall, A.: Programs TRAMO and SEATS, instructions for the users, Technical report, Dirección General de Análisis y Programación Presupuestaria, Ministerio de Economía y Hacienda, working paper 97001 (1998)Google Scholar
- 14.Hyndman, R.J., Khandakar, Y.: Automatic time series for forecasting: the forecast package for R. No. 6/07. Monash University, Department of Econometrics and Business Statistics (2007)Google Scholar
- 15.Hyndman, R.J.: Data from the M-competitions. Comprehensive R Archive Network. http://cran.r-project.org/web/packages/Mcomp/
- 19.McCleary, R., Hay, R.: Applied Time Series Analysis for the Social Sciences. Sage Publications, Beverly Hills (1980)Google Scholar
- 21.Ord, K., Lowe, S.: Automatic forecasting. Am. Stat. 50(1), 88–94 (1996)Google Scholar
- 23.R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org (2008)