@INPROCEEDINGS{lang09icml, author = "Lang, Tobias and Toussaint, Marc", title = "Approximate Inference for Planning in Stochastic Relational Worlds", booktitle = "Proc.\textasciitilde of the Int.\textasciitilde Conf.\textasciitilde on Machine Learning (ICML)", year = "2009", pages = "585--592", month = "June", abstract = "Relational world models that can be learned from experiencein stochastic domains have received significant attention recently. However,efficient planning using these models remains a major issue.We propose to convert learned noisy probabilisticrelational rules into a structured dynamic Bayesian networkrepresentation. Predicting the effects of action sequences using approximateinference allows for planning in complex worlds.We evaluate the effectiveness of our approach for online planning in a 3Dsimulated blocksworld with an articulated manipulator and realistic physics.Empirical results show that our method can solve problems where existing methodsfail.", folder = "Robotics", pdf = "http://www.user.tu-berlin.de/lang/pub/lang09icml.pdf", pdfurl = "http://www.user.tu-berlin.de/lang/pub/lang09icml.pdf" }