Original title:
Bayesian Entropic Evolution
Authors:
Gottvald, Aleš Document type: Papers Conference/Event: Mendel 2000 /6./ - Soft Computing, Brno (CZ), 2000-06-07 / 2000-06-09
Year:
2000
Language:
eng Abstract:
We develop a quantitative and experimantally testable theory of evolution, based on Bayesian and Entropic concepts. Probability and information are given central roles in evolutionary processes. Substantial evidence is now available that all logically consistent probabilistic transformations, in evolutionary processes or elsewhere, obey fundamental laws of Bayesian Probability Theory. In particular, Bayes' Theorem and the maximum Entropy Principle imply many quantitative and testable consequences also in evolutionary processes, in full analogy with Statistical Mechanics. Every evolutionary process may be treated as a chain of Bayesian probabilistic interfereces from incomplete information. Only systems equipped with a Bayesian processing layer react most rationally to a changing information environment, which brings some evolutionary advantages. Project no.: CEZ:AV0Z2065902 (CEP), ME 181 Funding provider: GA MŠk Host item entry: Mendel 2000 - 6th International Conference on Soft Computing, ISBN 80-214-1609-2 Note: Související webová stránka: mailto:gott@isibrno.cz
Institution: Institute of Scientific Instruments AS ČR
(web)
Document availability information: Fulltext is available at the institute of the Academy of Sciences. Original record: http://hdl.handle.net/11104/0100970