Page 183 - Contributed Paper Session (CPS) - Volume 6

CPS1867 Winita S. et al.
we get more flexible model since the functions can be adjusted to the
behaviour of the series.
In the next step, we apply NN to model the irregular component. NN is a
powerful method in handling the nonlinearity and uncertainty found in the
series. We observe several nodes in the hidden layer varying from 1 to 10, and
combine with a number of inputs that proportional to the period. NN with
the smallest RMSE and random residuals will be the chosen one.
Furthermore, results show that the hybrid SSA-NN(6-8-1) yields the best
performance in comparison with other methods in the mentioned literature.
Its MAE and MAPE are even smaller than those obtained by the hybrid method
based on local linear neuro-fuzzy model and optimized singular spectrum
analysis, named OSSA-LLNF (see Abdollahzade, Miranian, Hassani, &
Iranmanesh, 2015). The methodology discussed in this paper can be applied
in other cases.

References
1. Abdollahzade, M., Miranian, A., Hassani, H., & Iranmanesh, H. (2015). A
new hybrid enhanced local linear neuro-fuzzy model based on the
optimized singular spectrum analysis and its application for nonlinear and
chaotic time series forecasting. Information Sciences, 295, 107–125.
2. Adhikari, R., & Agrawal, R. K. (2012). Forecasting strong seasonal time
series with artificial neural networks. Journal of Scientific and Industrial
Research, 71(October 2012), 657–666.
3. Afshar, K., & Bigdeli, N. (2011). Data analysis and short term load
forecasting in Iran electricity market using singular spectral analysis (SSA).
Energy, 36(5), 2620–2627.
4. Brockwell, P. J., & Davis, R. A. (2002). Introduction to time series and
forecasting (2nd ed.). Springer-Verlag.
5. Elsner, J. B., & Tsonis, A. A. (1996). Singular Spectrum Analysis A New Tool
in Time Series Analysis. Springer Science & Business Media.
6. Golyandina, N. (2010). On the choice of parameters in singular spectrum
analysis and related subspace-based methods. Stat Interface, 3(3), 259–
279.
7. Golyandina, N., & Korobeynikov, A. (2014). Basic singular spectrum
analysis and forecasting with R. Computational Statistics & Data Analysis,
71, 934–954.
8. Golyandina, N., Nekrutkin, V., & Zhigljavsky, A. (2001). Analysis of Time
Series Structure: SSA and related techniques (Vol. 90). Chapman &
Hall/CRC, Boca Raton, FL.
9. Golyandina, N., & Zhigljavsky, A. (2013). Singular Spectrum Analysis for
time series. Springer Science & Business Media.

172 | I S I W S C 2 0 1 9