Time series modeling based on fuzzy analysis of position-binary components of historical data

R. Rzaev^a, Z. Jamalov^b, T. Mehdiyev^c, V. Hasanov<sup>c

a</sup> Institute of Cybernetics, National Academy of Sciences of Azerbaijan, Baku, Azerbaijan
^b Baku State University, Baku, Azerbaijan
^c Institute of Control Systems, National Academy of Sciences of Azerbaijan, Baku, Azerbaijan

Abstract: It is offered new predictive models of volatile time series based on fuzzy analysis of position-binary components of historical data. One of the distinctive features of the proposed models are rules fuzzification of historical data and defuzzification of fuzzy forecasts. In the context of this study, we propose a new criterion for assessing the adequacy of a model based on the use of the Hamming metric, which is on par with classical statistical evaluation criteria was used to evaluate the results.

Keywords: time series, fuzzy set, fuzzy prediction, fuzzy relation, point estimation, Hamming distance.

UDC: 629.4.067

Received: 10.01.2015
Revised: 29.05.2015

Bibliographic databases:

•