Network information and connected correlations
Network information and connected correlations
复制标题
DOI：
10.1103/physrevlett.91.238701
复制
发表时间：
2003-12-05
影响因子：
8.6
通讯作者：
Bialek, W
中科院分区：
文献类型：
--
作者：
Schneidman, E;Still, S;Bialek, W
熵和信息提供了网络中元素之间相关性的自然度量。我们在这里构建了连接相关函数的信息论模拟：不可约N点相关性是通过N个变量的联合分布相对于所有观察到的N-1个变量分布所允许的最大熵的熵减少来测量的。我们计算的“连接信息”的条款的几个例子，并表明，它也使分解的信息，是由人口的元素有关的外部来源。
Entropy and information provide natural measures of correlation among elements in a network. We construct here the information theoretic analog of connected correlation functions: irreducible N-point correlation is measured by a decrease in entropy for the joint distribution of N variables relative to the maximum entropy allowed by all the observed N-1 variable distributions. We calculate the "connected information" terms for several examples and show that it also enables the decomposition of the information that is carried by a population of elements about an outside source.