Inference and Learning with Model Uncertainty in Probabilistic Logic Programs

An issue that has so far received only limited attention in probabilistic logic programming (PLP) is the modelling of so-called epistemic uncertainty, uncertainty about model itself. Accurately quantifying this paramount to robust inference, learning and ultimately decision making. We introduce BetaProbLog, a PLP language can uncertainty. BetaProbLog sound semantics, an effective inference algorithm combines Monte Carlo techniques with knowledge compilation, parameter algorithm. empirically outperform state-of-the-art methods on tasks second-order Bayesian networks, digit classification discriminative presence