|
Time series can be used by the specialists in many sciences, not only mathematicians, physicists and computer scientists, recently, also specialists in the field of finance to modelling any processes and phenomena which involves temporal measurement e.g. weather forecasting, earthquake prediction, power demand forecasting or exchange rate prediction. A concept of the fractal dimension as a measure of variability of a time-series graph of exchange rates is presented in this paper. Two methods of calculating the fractal dimension of time-series graph are described. The first method is the most known box method by Kolmogorov, the second one is a new method that uses the Hölder exponent. The paper contains the results of computermodelling the values of analysed exchange rates of some currencies with use the fractal dimensions obtained in above-mentioned methods in 2000-2016 period.
|
