Belief χ2 Divergence-Based Dynamical Complexity Analysis for Biological Systems
Published in International Conference on Machine Learning for Cyber Security, 2023
Abstract: Physiological signals contain the information of physical state in healthy systems. Especially, the complexity of dynamical time series data is a valid indicator to measure the pathological states. However, how to quantify the complexity of physical signals is still an open issue. In this paper, a novel complexity analysis algorithm based on divergence, called DCA, is proposed to figure out the complexity of biological system time series. Specifically, DCA algorithm splits biological systems time series data into different slices with boundaries. In addition, the feature of each time series data is extracted by converted into the basic probability assignments (BPAs) based on the Dempster-Shafer (D-S) evidence theory. DCA algorithm considers that the average divergence of BPAs indicates the complexity in a piece of time series. Moreover, an application in cardiac inter-beat interval time series is carried out to demonstrate the effectiveness of the proposed algorithm, which performs well in a pathological states analysis issue.
Recommended citation:Zhang, L., Xiao, F. (2023). Belief χ2 Divergence-Based Dynamical Complexity Analysis for Biological Systems. In: Xu, Y., Yan, H., Teng, H., Cai, J., Li, J. (eds) Machine Learning for Cyber Security. ML4CS 2022. Lecture Notes in Computer Science, vol 13657. Springer, Cham. https://doi.org/10.1007/978-3-031-20102-8_13