Location Weight of GSTAR Model for High Variability of Rainfall Data

  • Abstract
  • Keywords
  • References
  • PDF
  • Abstract

    The use of location weights in the spatio temporal models took part in the accuracy of the model. The weights that often used are uniform weight, inverse distance weight, and normalized cross-correlation weight. These location weights consider the closeness between locations. For data that have high degree of variability, the use of these location weights becomes less relevant. Therefore, it is necessary to consider variability aspects of observational data as the location weights, that is the normalize ratio variance weight. The study was conducted to develop GSTAR-SUR model with normalize ratio variance weight and its application on rainfall data. The data used is ten daily rainfall data in the region Blimbing, Singosari, Karangploso, Dau, and Wagir. Based on the results of the study, indicated that the GSTAR-SUR ((1) (1,2,3,12,36) model that used ratio variance as location weight are more accurately to forecast the rainfall data that has high variability and extreme point.



  • Keywords

    GSTAR-SUR, rainfall, ratio variance

  • References

      [1] Borovkova, S.A., Lopuha, H.P. and Ruchjana, B.N. 2002. Generalized S-TAR with Random Weights. Proceeding of the 17th International Workshop on Statistical Modeling. Chania-Greece.

      [2] Davidson, Russell and MacKinnon, James G. 1993. Estimation and Inference in Economics. Oxford University Press. UK.

      [3] Greene, W.H. 2002. Econometric Analysis (5thed). Prentice Hall.

      [4] Iriany, A., Suhariningsih, Ruchjana, B.N., and Setiawan. 2013. Prediction of Precipitation Data at Batu Town using The GSTAR(1,p)-SUR. Journal of Basic and Applied Scientific Research.3(6): vol 3 no 6.

      [5] Nainggolan, Nelson. 2012. Development of GSTAR Model with Assumption ARCH and its Application to Inflation. Dissertation Unpublihed. Bandung: ITB.

      [6] Pfeifer, P.E and Deutsch, S.J. 1980a. Stationarity and Invertibility Regions for Low Order STARMA Models. Communications in Statistics-Simulation and Computation 9 (5). P. 551-562.

      [7] Pfeifer, P.E and Deutsch, S.J.. 1980b. Identification and Interpretation of First Orde Space-Time ARMA Models.Technometrics, 22 (1), 397-408.

      [8] Ruchjana, B.N. 2001. Study on the Weight Matrix in the Space-Time Autoregressive Model. Proceeding of the Tenth International Symposium on Applied Stochastic Models and Data Analysis (ASMDA). Gerard Govaert, et al. Editor. Universite de Technologie Compiegne. Prancis. Vol. 2/2. p. 789-794.

      [9] Ruchjana, B.N. . 2002.A Generalized Space Time Autoregressive Model and its Application to Oil Production Data. Dissertation Unpublihed. Bandung : ITB.

      [10] Suhartono dan Atok, R.M. 2006. Pemilihan Bobot Lokasi yang Optimal pada Model GSTAR. Prosiding Konferensi Nasional Matematika XIII, Universitas Negeri Semarang.

      [11] Suhartono and Subanar. 2006. The Optimal Determination of Space Weight in GSTAR Model by Using Cross-sorrelation Inference Journal of Quantitative Methods: Journal Devoted the Mathematical and Statistical Application in Various Field.

      [12] Setiawan. 1992. Study on Seemingly Unrelated Regression (SUR) and its application to Almost Ideal Demand System (AIDS). Thesis Unpublished. Bogor : IPB.

      [13] Srivastava, V.K. and Giles, D.E.A. (1987). Seemingly unrelated regression equations models: estimation and inference. New York: Marcel Dekker. ISBN 978-0-82-477610-7.

      [14] Wutsqa, D.U dan Suhartono. 2010. Peramalan Deret Waktu Multivariat Seasonal pada Data Pariwisata dengan VAR-GSTAR. Jurnal Ilmu Dasar Vol. II no. 1.

      [15] Zellner, A. (1962). An efficient method of estimating seemingly unrelated regression equations and tests for aggregation bias. Journal of the American Statistical Association 57: 348–368.




Article ID: 26393
DOI: 10.14419/ijet.v8i1.9.26393

Copyright © 2012-2015 Science Publishing Corporation Inc. All rights reserved.