Using Cumulant Analysis For Entropic Complexity Measures

Abstract

A simplified approach based on the cumulant analysis of the Shannon entropy is proposed for measuring complexity. We report the result of such analysis for some generic problems (mixing r − component ideal gas, the simplest geometric ornament, Schrödinger's cat states, and the logistic map). We argue that the new measures have benefits compared to the currently practiced measures in the Shiner-Davison-Landsberg and Lopes-Ruiz-Mancini-Calbet approaches.