Interaction Information for Causal Inference: The Case of Directed Triangle

被引：0
|
作者：
Ghassami, AmirEmad [1 ]
Kiyavash, Negar [1 ,2 ]
机构：
[1] Univ Illinois, Coordinated Sci Lab, Dept ECE, Urbana, IL 61801 USA
[2] Univ Illinois, Coordinated Sci Lab, Dept ISE, Urbana, IL USA
来源：
2017 IEEE INTERNATIONAL SYMPOSIUM ON INFORMATION THEORY (ISIT) | 2017年
关键词：
Mutual information generalization; Interaction information; Causal inference;
D O I：
暂无
中图分类号：
TP301 [理论、方法];
学科分类号：
081202 ;
摘要：
Interaction information is one of the multivariate generalizations of mutual information, which expresses the amount of information shared among a set of variables, beyond the information shared in any proper subset of those variables. Unlike (conditional) mutual information, which is always non negative, interaction information can be negative. We utilize this property to find the direction of causal influences among variables in a triangle topology under some mild assumptions.
引用
收藏
页码：1326 / 1330
页数：5
相关论文
共 50 条
12345