On a geometric mean and power-law statistical distributions
arXiv:cond-mat/0507414
Abstract
For a large class of statistical systems a geometric mean value of the observables is constrained. These observables are characterized by a power-law statistical distribution.