skip to main content

PERAMALAN PADA RUNTUN WAKTU DENGAN POLA TREND MENGGUNAKAN SSA-LRF

*Diah Safitri  -  Departemen Matematika, Universitas Gadjah Mada, Indonesia
Gunardi Gunardi  -  Departemen Matematika, Universitas Gadjah Mada, Indonesia
Nanang Susyanto  -  Departemen Matematika, Universitas Gadjah Mada, Indonesia
winita Sulandari  -  Program Studi Statistika, Universitas Sebelas Maret, Indonesia
Open Access Copyright 2023 Jurnal Gaussian under http://creativecommons.org/licenses/by-nc-sa/4.0.

Citation Format:
Abstract

Singular Spectrum Analysis-Linear Recurrent Formulae (SSA-LRF) is a forecasting method that starts by decomposing time series data into several independent and interpretable components. SSA-LRF does not have any assumptions that must be fulfilled thus it is more flexible to use. In this research, an empirical study of time series forecasting that has a trend data pattern will be carried out using SSA-LRF without difference transformation and with difference transformation. A difference transformation is performed because the data has a trend pattern. Although there are no assumptions that must be met in forecasting using SSA-LRF, it is expected that difference transformation will produce better forecasting accuracy than without difference transformation process. There are three data used in this research. The first is data from Wei's book (2006), this data is called series W8 and is a simulation data. The second data is the number of railway passengers in the Java region. The third data is Mauna Loa atmospheric CO2 concentration data obtained from R software. Forecasting using SSA-LRF without difference transformation and with difference transformation on all three data resulted in accurate forecasting values, and difference transformation improved the accuracy values

Fulltext View|Download
Keywords: SSA; SSA-LRF; Trend;Time Series

Article Metrics:

  1. Elsner, J. B. dan A. A. Tsonis. 1996. Singular Spectrum Analysis A New Tool in Time Series Analysis. Springer Science & Business Media
  2. Golyandina, N. dan A. Korobeynikov. 2014 Basic singular spectrum analysis and forecasting with R. Computational Statistics & Data Analysis. 71:934-954
  3. Golyandina, N. dan A. Zhigljavsky. 2013. Singular Spectrum Analysis for Time Series. Springer
  4. Golyandina, N., V. Nekrutkin, dan A. A. Zhigljavsky. 2001. Analysis of Time Series Structure : SSA and Related Techniques, Chapman & Hall/CRC
  5. Hassani, H. 2007. Singular spectrum analysis: methodology and comparison. Journal of Data Science, 5:239-257
  6. Hassani, H., A. S. Soofi, dan A. A. Zhigljavsky. 2010. Predicting daily exchange rate with singular spectrum analysis. Nonlinear Analysis:Real World Applications, 11: 2023-2034
  7. Hassani, H., S. Heravi, dan A. Zhigljavsky. 2009. Forecasting European industrial production with singular spectrum analysis. International Journal of Forecasting, 25:103-118
  8. Mahmoudvand, R., P. C. Rodrigues, dan M. Yarmohammadi. 2019. Forecasting daily exchange rates: A comparison between SSA and MSSA. Statistical Journal, 17 (4):599-616
  9. Makridakis, S. G., S. C. Wheelwright, dan R. J. Hyndman. 2008. Forecasting Methods and Applications. Wiley
  10. Wei, W. W. S. 2006. Time Series Analysis Univariate and Multivariate Methods, second edition, Pearson Education Inc

Last update:

No citation recorded.

Last update:

No citation recorded.