Learning with Bayesian networks and probability trees to approximate a joint distribution | IEEE Conference Publication | IEEE Xplore