|
Abstract
|
In this article, we investigate the visibility algorithm which has been introduced to map a time series into a network. At first, with an example, we show that the visibility algorithm cannot be used to extract the correlation information of a series. Further, using the horizontal visibility algorithm, we demonstrate that the degree distributions of the fractal Brownian motions can well be fitted with a nonlinear exponential functions. In addition, the correlation between the series data and the degree sequence is calculated for the Hurst index prediction purpose. The obtained results are in contrast with some recent studies.
|