NONLINEAR PHENOMENA IN COMPLEX SYSTEMS
An Interdisciplinary Journal

2006, Vol.9, No.3, pp.276-282


Empirical Studies of Random Multiplicative Stochastic Processes: Revisit to Lognormals.
N. Kobayashi, Y. Sasaki, O. Moriyama, S. Matsushita and M. Matsushita

The statistics involving random multiplicative stochastic processes have been studied empirically. Examples taken here are the duration distribution of disability for aged people, the life span of animals and the population distribution of prefectures in Post-World War II Japan. We found that lognormal distributions show excellent fit with various data for the duration distribution of disability and the life span of animals. The good data fitting for both cases by lognormal distributions indicates that the incidence of both events can be considered as many independent subprocesses in succession. We also found that a two-lognormals distribution fits very well over the entire region for the population distribution of prefectures in post-war Japan. This result implies that in this case the size segregation is also relevant, in addition to the random multiplicative stochastic processes.
Key words: lognormal distribution, duration of disability, life span of animals, population distribution, two-lognormals distribution

Full text:  Acrobat PDF  (933KB)  



ContentsJournal Home Page

Copyright © Nonlinear Phenomena in Complex Systems. Last updated: September 6, 2006