A Characterization of Interventional Distributions in Semi-Markovian Causal Models

We offer a complete characterization of the set of distributions that could be induced by local interventions on variables governed by a causal Bayesian network of unknown structure, in which some of the variables remain unmeasured. We show that such distributions are constrained by a simply formulated set of inequalities, from which bounds can be derived on causal effects that are not directly measured in randomized experiments.

• Type: Conference
• Year: 2006
• Where: AAAI
• Authors: Jin Tian, Changsung Kang, Judea Pearl
• Keywords: AAAI 2006, Causal Bayesian Network, Complete Characterization, Intelligent Agents, Local Interventions