PC algorithm for complex survey data via resampling

The PC algorithm is one of the main methods for learning the structure of a Bayesian network from sample data. The algorithm uses conditional independence tests for model selection in graphical modeling and it is based on assumption of independent and identically distributed observations (i.i.d). The i.i.d. assumption is almost never valid for sample surveys data since most of the commonly used survey designs employ stratification and/or cluster sampling and/or unequal selection probabilities. The impact of complex design on i.i.d. based procedures can be very severe leading to erroneous results, then alternative procedures are needed which allow for complex designs. The aim is to modify the PC algorithm using resampling methods for finite population in order to take into account the complexity of sampling design in the learning process.