Conjunctive combined causal rules mining

Discovering causal relationships among a set of observed variables is a very important and essential problem in science. In many fields, predicting causes can help to avoid harmful consequences. Learning Bayesian network (BN), and Randomized Controlled Trials (RCTs) play a major role in Causal discovery. Existing algorithms fail to discover causal relationships on non-fixed structures, the cost of these algorithms is very high, and they are employed to discover only single cause rules from certain data. In this paper, we are interested in reducing the cost of Causal discovery by employing the study of frequent itemsets mining to discover conjunctive combined Causal rules from uncertain data. We propose an algorithm called CCCRUD for this problem and evaluate it on real datasets. We believe that this is the first work that address the problem of discovering casual rules in the context of uncertain data and conjunctive targets.