Modified generalized sample entropy and surrogate data analysis for stock markets

In this paper a modified method of generalized sample entropy and surrogate data analysis is proposed as a new measure to assess the complexity of a complex dynamical system such as stock market. The method based on Hausdorff distance presents a different way of time series patterns match showing distinct behaviors of complexity. Simulations are conducted over synthetic and real-world data for providing the comparative study. Results show that the modified method is more sensitive to the change of dynamics and has richer information. In addition, exponential functions can be used to successfully fit the curves obtained from the modified method and quantify the changes of complexity for stock market data. (C) 2015 Elsevier B.V. All rights reserved.