Complexity Methods Used in the Study of Some Real Systems with Weak Characteristic Properties

摘要

Tsallis statistical analysis and other possible use of q-functions are considered here in the study of some real systems where the observed data are scarce or a typical characteristic property is not significant. The analysis is performed for some systems concerning opinion formation, hydrologic data, financial series etc. Here, Tsallis q-triplet is used to assist the identification of the most appropriate distribution, and in another case we consider q-entropy optimization in the study of mixed states of the system. Next, we propose to extend the idea of the distance from the stationary state indicated by q-stationary parameter in Tsallis statistics, assuming that the distance from a proper regime could be expressed by a corresponding q-function. Therefore an ad hoc q-log-periodic form is introduced to describe the near to log-periodic behavior.