W. Rogoza, G. Ishchenko
Èlektron. model. 2022, 44(3):42-49
https://doi.org/10.15407/emodel.44.03.042
ABSTRACT
A comparative analysis of two deterministic methods of short-term forecasting of time series in the conditions of a limited amount of experimental data is performed. The first method is based on the construction of so-called partial predictive models in the form of Kolmogorov-Gabor second-order polynomials. In the second method, in addition to these polynomials, predictive partial models are built in the form of increments of changes in the parameters of the object under study. This makes it possible to increase the amount of data for training forecast models. Due to this increase, the researcher gets more accurate predictions for each parameter. A comparison of the results of the forecast on the example confirms this conclusion.
KEYWORDS
time series, deterministic methods, comparative analysis of forecasting methods.
REFERENCES
- Box, G. and Jenkins, G. (1970), Time Series Analysis: Forecasting and Control, Holden-Day, San Francisco, USA.
- Kalman, R.E. (1960), “A new approach to linear filtering and prediction problems”, Journal of Basic Engineering, Vol. 82, no. 1, pp. 35-45.
https://doi.org/10.1115/1.3662552 - Ivakhnenko, A.G. (1982), Inductive method of self-organization of complex systems, Naukova Dumka, Kyiv, USSR.
- Rogoza, W.S. and Ischenko, G.V. (2022), “Method for forecasting short-term series using sensitivity functions”, Elektronne modelyuvannya, Vol. 44, no. 1, pp. 29-42.
https://doi.org/10.15407/emodel.44.01.029