arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1208.2955cs.CC
收藏

随机分布,随机性,相依性

Enumerable Distributions, Randomness, Dependence

Leonid A. Levin

摘要:

无限序列(以及它们的有限前缀)中的互信息I在许多情况的理论分析中是必不可少的。然而,长期以来,它的正确定义一直难以捉摸。我通过将Kolmogorov复杂性理论从测度推广到半测度,即测度集的下确界来解决这个问题。作为凹泛函而不是线性泛函,半测度处理起来相当微妙。然而,他们充分掌握各种理论和实践知识。信息的一个简单的下界i $(α:β) = \sup\,_{x\in N}\,(K(x) - K(x|α) - K(x|β)) $对于Martin-Lof随机$ α,β$是紧的。对于所有序列I $(α:β) $,其特征在于i $(α':β') $在具有$ U(α')=α, U(β')=β$的随机$ α',β' $上最小。

英文摘要:

Mutual information I in infinite sequences (and in their finite prefixes) is essential in theoretical analysis of many situations. Yet its right definition has been elusive for a long time. I address it by generalizing Kolmogorov Complexity theory from measures to SEMImeasures i.e, infimums of sets of measures. Being concave rather than linear functionals, semimeasures are quite delicate to handle. Yet, they adequately grasp various theoretical and practical scenaria. A simple lower bound i$(α:β) = \sup\,_{x\in N}\,(K(x) - K(x|α) - K(x|β)) $ for information turns out tight for Martin-Lof random $ α,β$. For all sequences I$(α:β) $ is characterized by the minimum of i$(α':β') $ over random $ α',β' $ with $ U(α')=α, U(β')=β$.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑