Title :
An Improved K-Line Fractal Dimension and Its Application on Stock Time Series Segmentation
Author :
Ni, Liping ; Ni, Zhiwei ; Gao, Yazhuo
Author_Institution :
Key Lab. of Process Optimization & Intell. Decision-making, Hefei Univ. of Technol., Hefei, China
Abstract :
Fractal dimension can be used to analyze stock time series. This paper proposes a new K-line fractal dimension which is calculated from the positive or negative line separately. Accordingly, we call it positive line fractal dimension or negative line fractal dimension. Then negative line fractal dimension is used to segment time series. Experiment results show this new K-line fractal dimension can reflect the uptrend or downtrend of stock market better than traditional K-line fractal dimension.
Keywords :
fractals; stock markets; time series; improved k-line fractal dimension; stock market; stock time series segmentation; Equations; Estimation; Fractals; Indexes; Mathematical model; Stock markets; Time series analysis; K-line; negative line dimension; positive line fractal dimension; segmentation;
Conference_Titel :
Business Intelligence and Financial Engineering (BIFE), 2010 Third International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-7575-9
DOI :
10.1109/BIFE.2010.83
