Applications of Multifractal Detrended Fluctuation Analysis
Year: 2023
Author: Dongxu Dai, Yan Hu, Hao Qian, Guoqiang Qi, Renchang Sun, Jian Wang, Yan Wang, Hailong Lang, Minggang Li
Journal of Information and Computing Science, Vol. 18 (2023), Iss. 2 : pp. 81–100
Abstract
In recent years, multifractal detrended fluctuation analysis (MF-DFA) has become an important tool for detecting the scale and long correlation of non-stationary time series. With the continuous development of multifractal theory, researchers have widely applied it in physics, chemistry, biology, economy, etc. In this paper, we briefly review various applications of MF-DFA, and present some empirical research using one- and two-dimensional (2D) MF-DFA. 1D MF-DFA is always applied in financial markets, energy markets, heartbeat, and atmospheric science. Furthermore, 2D MF-DFA has been studied in surface science such as image segmentation, medical image classification. In this paper, we use 1D MF-DFA to explore the market efficiency of Korean stock market, and adopt 2D MF-DFA to segment images such as the license plate and hepatic cell image. In addition, we apply the proposed algorithm to segment transmission lines under icing condition, and the proposed method achieves satisfactory segmentation results.
Journal Article Details
Publisher Name: Global Science Press
Language: English
DOI: https://doi.org/10.4208/JICS-2023-006
Journal of Information and Computing Science, Vol. 18 (2023), Iss. 2 : pp. 81–100
Published online: 2023-01
AMS Subject Headings:
Copyright: COPYRIGHT: © Global Science Press
Pages: 20
Keywords: MF-DFA Hurst exponent Feature extraction Image segmentation.
Author Details
Dongxu Dai Email
Yan Hu Email
Hao Qian Email
Guoqiang Qi Email
Renchang Sun Email
Jian Wang Email
Yan Wang Email
Hailong Lang Email
Minggang Li Email