First-order probabilistic inference

By David Poole. International Joint Conference on Artificial Intelligence, 2003

531 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/3d4606a66c81cf3a5fa647d80c5895e2301db4a9

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Practical solution techniques for first-order MDPs (2009, methodology): “Ostensibly, relation elimination and the technique of first-order variable elimination (FOVE) (Poole, 2003; de Salvo Braz et al., 2005; de Salvo Braz et al., 2006) appear similar since they both deal with lifted versions of variable elimination.”
Counting Belief Propagation (2009, methodology): “First-order probabilistic models essentially combine graphical models with elements of firstorder logic by defining template factors (such as Poole’s parfactors [13]) that apply to whole sets of objects at once.”
Statistical Relational Artificial Intelligence: Logic, Probability, and Computation (2016, methodology): “One central question is that of lifted inference [Poole, 2003], that is, whether one can perform probabilistic inference without grounding out the probabilistic relational model first.”
MEBN: A language for first-order Bayesian knowledge bases (2008, background): “Recent years have seen a rapid evolution of increasingly powerful languages for computational probabilistic reasoning (e.g., Buntine, 1994; D’Ambrosio, et al, 2001; Getoor et al, 2001; Gilks et al, 1994; Glesner and Koller, 1995; Heckerman, et al., 2004; Jaeger, 2001; Kersting and De Raedt, 2001; Ko”
Probabilistic models with unknown objects (2006, background): “Lifted inference algorithms that exploit generalization across objects [Pfeffer et al., 1999; Poole, 2003; de Salvo Braz et al., 2005] also apply naturally to BLOG, and might yield dramatic speed-ups on some problems.”
Markov Chains on Orbits of Permutation Groups (2012, background): “Poole introduced the notion of lifted probabilistic inference as a variation of variable elimination taking advantage of the symmetries in graphical models resulting from probabilistic relational formalisms [36].”
Lifted First-Order Probabilistic Inference (2005, methodology): “We use the implementation available at http://l2r.cs.uiuc.edu/˜cogcomp to compare average run times between lifted and propositional inference (which produce the exact same results) for two different models while increasing the number of objects in the domain.”
Relational Logistic Regression: The Directed Analog of Markov Logic Networks (2014, background): “All rights reserved. crime (which depends on how many other people could have committed the crime) (Poole 2003) we could consider the population to be the population of the neighbourhood, the population of the city, the population of the country, or the population of the whole world.”
Reasoning about Probabilities in Dynamic Systems using Goal Regression (2013, background): “We do not attempt to do a full survey here, but this has received a lot of attention, often under reasonable assumptions such as the ability to factorize domains (Poole, 2003; Gogate and Domingos, 2010).”
Lifted Rule Injection for Relation Embeddings (2016, methodology): “In statistical relational learning this type of approach is often referred to as lifted inference or learning (Poole, 2003; Braz, 2007) because it deals with groups of random variables at a first-order level.”
A First-Order Logic of Probability and Only Knowing in Unbounded Domains (2016, background): “Recently in AI, limited versions of probabilistic logics have been discussed (Poole 2003; Domingos and Webb 2012), with things like a ﬁnite domain assumption built-in.”
Open-Universe Weighted Model Counting (2017, background): “…propositional resolution steps, there are a number of extensions to WMC that leverage the first-order structure of template models by lumping together classes of objects (Poole 2003; de Salvo Braz, Amir, and Roth 2005; Gogate and Domingos 2011; Beame et al. 2015), referred to as lifted inference.”
Reasoning about Probabilities in Unbounded First-Order Dynamical Domains (2017, background): “Limited versions of probabilistic logics have also become popular in the machine learning literature [Poole, 2003; Domingos and Webb, 2012; Getoor and Taskar, 2007], with things like a finite domain assumption built-in.”
Probabilistic (logic) programming concepts (2015, methodology): “…CLP( BN ) essentially replace the speciﬁc random variables in the CPTs of Bayesian networks by logically deﬁned random variable templates, commonly referred to as parameterized random variables or par-RVs for short (Poole, 2003), though the actual syntax amongst these systems diﬀers signiﬁcantly.”
Unifying logic and probability (2015, unknown): “Finally, the rapidly advancing techniques of lifted inference [28] aim to unify probability and logic at the inferential level, borrowing and generalizing ideas from logical theorem proving.”
Markov logic networks (2006, background): “Counting the number of true groundings of a first-order clause in a database is #P-complete in the length of the clause.”