Lecture on Stochastic Simulation Methods for Probabilistic Inference (2008)
Technique for approximate inference in Bayesian networks • Run repeated simulations of the world described by the network • Estimate the probabilities we are interested in by counting the...
Brian Milch, S. Russel, P. Norvig, Artificial Intelligence, A Modern
• Fundamental task: given observations, make inferences about initially unknown objects • But most RPM languages assume set of objects is fixed and known (Herbrand models) • Bayesian logic...