Open AccessMeasuring the Complexity of a Physiological Time Series: a Review
Author(s) -
Kazimieras Pukėnas,
Jonas Poderys,
Remigijus Gulbinas
Publication year - 2018
Publication title -
baltic journal of sport and health sciences
Language(s) - English
Resource type - Journals
eISSN - 2538-8347
pISSN - 2351-6496
DOI - 10.33607/bjshs.v1i84.299
Subject(s) - sample entropy , estimator , entropy (arrow of time) , computer science , mathematics , time series , artificial intelligence , algorithm , statistical physics , machine learning , statistics , physics , quantum mechanics
Research background and hypothesis. Complex Systems Theory indeed is a solid basis for a scientific approach in the analysis of living, learning, and evolving systems. A number of different entropy estimators have been applied to physiological time series attempting to quantify its complexity. Research aim. The aim of the paper is to review most popular complexity estimators (entropies) applied in biological, medical, sport and exercise sciences and their performances.Research results. Various measures of complexity were developed by scientists to compare time series and distinguish regular (e. g. periodic), chaotic, and random behavior. In this paper a brief review of most popular complexity estimators – Sample Entropy, Control Entropy, Spectral Entropy, Wavelet Entropy, Singular-Value Decomposition Entropy, Permutation Entropy, Base-Scale Entropy, Entropy based on Lempel-Ziv algorithm – and their performances is presented. In biological applications they are used to distinguish peculiarities in behavior of biological systems or may serve as non-invasive, objective means of determining physiological changes under steady or non-steady state conditions.Discussion and conclusions. The choice of a particular entropy estimator is determined by the goal type, the capability of estimators in characterizing the constraints on a physiological time series, its robustness to noise considering the above-mentioned advantages and disadvantages of particular algorithms. It is difficult to apply analytical solutions in the analysis of behavior of living, learning, and evolving systems and new approaches and solutions remain on the agenda.Keywords: physiological time series, complexity, entropy.