Time series forecasting by using a neural arima model based on wavelet decomposition
Article Sidebar
Main Article Content
Abstract
In the prediction of (stochastic) time series, it has been common to suppose that an individual predictive method – for instance, an Auto-Regressive Integrated Moving Average (ARIMA) model – produces residuals like a white noise process. However, mainly due to the structures of auto-dependence not mapped by a given individual predictive method, this assumption may easily be violated, in practice, as pointed out in Firmino et al. (2015). In order to correct it (and accordingly to produce more forecasts with more accuracy power), this paper puts forward a Wavelet Hybrid Forecaster (WHF) that integrates the following numerical techniques: wavelet decomposition; ARIMA models; Artificial Neural Networks (ANNs); and linear combination of forecasts. Basically, the proposed WHF can map simultaneously linear – by means of a linear combination of ARIMA forecasts – and non-linear – through a linear combination of ANN forecasts – auto-dependence structures exhibited by a given time series. Differently of other hybrid methodologies existing in literature, the WHF forecasts are produced carrying into account implicitly the information from the frequency presenting in the underlying time series by means of the Wavelet Components (WCs) obtained by the wavelet decomposition approach. All numerical results show that WHF method has achieved remarkable accuracy gains, when comparing with other competitive forecasting methods already published in specialized literature, in the prediction of a well-known annual time series of sunspot.
Downloads
Article Details
1. Proposal of Policy for Free Access Periodics
Authors whom publish in this magazine should agree to the following terms:
a. Authors should keep the copyrights and grant to the magazine the right of the first publication, with the work simultaneously permitted under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 that allows the sharing of the work with recognition of the authorship of the work and initial publication in this magazine.
b. Authors should have authorization for assuming additional contracts separately, for non-exclusive distribution of the version of the work published in this magazine (e.g.: to publish in an institutional repository or as book chapter), with recognition of authorship and initial publication in this magazine.
c. Authors should have permission and should be stimulated to publish and to distribute its work online (e.g.: in institutional repositories or its personal page) to any point before or during the publishing process, since this can generate productive alterations, as well as increasing the impact and the citation of the published work (See The Effect of Free Access).
Proposal of Policy for Periodic that offer Postponed Free Access
Authors whom publish in this magazine should agree to the following terms:
a. Authors should keep the copyrights and grant to the magazine the right of the first publication, with the work simultaneously permitted under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 [SPECIFY TIME HERE] after the publication, allowing the sharing of the work with recognition of the authorship of the work and initial publication in this magazine.
b. Authors should have authorization for assuming additional contracts separately, for non-exclusive distribution of the version of the work published in this magazine (e.g.: to publish in institutional repository or as book chapter), with recognition of authorship and initial publication in this magazine.
c. Authors should have permission and should be stimulated to publish and to distribute its work online (e.g.: in institutional repositories or its personal page) to any point before or during the publishing process, since this can generate productive alterations, as well as increasing the impact and the citation of the published work (See The Effect of Free Access).
d. They allow some kind of open dissemination. Authors can disseminate their articles in open access, but with specific conditions imposed by the editor that are related to:
Version of the article that can be deposited in the repository:
Pre-print: before being reviewed by pairs.
Post-print: once reviewed by pairs, which can be:
The version of the author that has been accepted for publication.
The editor's version, that is, the article published in the magazine.
At which point the article can be made accessible in an open manner: before it is published in the magazine, immediately afterwards or if a period of seizure is required, which can range from six months to several years.
Where to leave open: on the author's personal web page, only departmental websites, the repository of the institution, the file of the research funding agency, among others.
References
ADHIKARI, R.; AGRAWAL, R. K. (2013) A Homogeneous Ensemble of Artificial Neural Networks for Time Series Forecasting. International Journal of Computer Applications, v. 32, n. 7, p. 8.
BATES, J. M.; GRANGER, C. W. J. (1969) The Combination of ForecastsJournal of the Operational Research Society.
DONOHO, D. L.; JOHNSTONE, J. (1994) M. Ideal spatial adaptation by wavelet shrinkage. Biometrika, v. 81, n. 3, p. 425–455.
FIRMINO, P. R. A.; DE MATTOS NETO, P. S. G.; FERREIRA, T. A. (2015) E. Error modeling approach to improve time series forecasters. Neurocomputing, v. 153, p. 242–254.
HAMILTON, J. D. (1994) Time Series Analysis,1ed. New Jersey : Princeton University Press.
HAVEN, E.; LIU, X.; SHEN, L. (2012) De-noising option prices with the wavelet method. European Journal of Operational Research, v. 222, n. 1, p. 104–112.
HAYKIN, S. S. (2001) Redes Neurais, 2ed. Porto Alegre: Bookman.
KHASHEI, M.; BIJARI, M. A. (2011) New Hybrid Methodology for Nonlinear Time Series Forecasting. Modelling and Simulation in Engineering, v. 2011, p. 1–5.
KUBRUSLY, C. S. (2011) The Elements of Operator Theory, 2 ed. New York: Birkhäuser.
KUBRUSLY, C. S.; LEVAN, N. (2006) Abstract wavelet generated by hilbert space shift operators. Adavances in mathematical Sciences and applications, v. 16, p. 643–660.
LEVAN, N.; KUBRUSLY, C. S. (2003) A wavelet “time-shift-detail” decomposition. Mathematics and Computers in Simulation, v. 63, n. 2, p. 73–78.
LIU, L.-M. (2006) Time Series Analysis and Forecasting. second ed. Chicago, IL: Scientific Computing Associates Corporation.
LUTKEPOHL, H.( 2006) Forecasting with VARMA Models. In: Handbook of Economic Forecasting. [s.l.] Elsevier, v. 1, p. 287–325.
MALLAT, S. (2009) A Wavelet Tour of Signal Processing: The Sparse Way, 3 ed. Burlington: Elsevier Inc.
RAGSDALE, C. (2004) Spreadsheet Modeling & Decision Analysis: A Practical Introduction to Management Science. Fourth edi ed. [s.l.] South-Western.
TEIXEIRA JR, L. A. et al. (2015) Artificial Neural Network and Wavelet decomposition in the Forecast of Global Horizontal Solar Radiation. Sobrapo, v. 35, n. 1, p. 1–16.
ZHANG, G. P. (2003) Time series forecasting using a hybrid ARIMA and neural network model. Neurocomputing, v. 50, p. 159–175.