Efficient Weighted Model Integration via SMT-Based Predicate Abstraction

Weighted model integration (WMI) is a recent formalism generalizing weighted model counting (WMC) to run probabilistic inference over hybrid domains, characterized by both discrete and continuous variables and relationships between them. Albeit powerful, the original formulation of WMI suffers from some theoretical limitations, and it is computationally very demanding as it requires to explicitly enumerate all possible models to be integrated over. In this paper we present a novel general notion of WMI, which fixes the theoretical limitations and allows for exploiting the power of SMTbased predicate abstraction techniques. A novel algorithm combines a strong reduction in the number of models to be integrated over with their efficient enumeration. Experimental results on synthetic and real-world data show drastic computational improvements over the original WMI formulation as well as existing alternatives for hybrid inference.

[1]  Vibhav Gogate,et al.  Approximate Inference Algorithms for Hybrid Bayesian Networks with Discrete Constraints , 2005, UAI.

[2]  John Eccleston,et al.  Statistics and Computing , 2006 .

[3]  Nils J. Nilsson,et al.  Artificial Intelligence , 1974, IFIP Congress.

[4]  Marco Roveri,et al.  Computing Predicate Abstractions by Integrating BDDs and SMT Solvers , 2007, Formal Methods in Computer Aided Design (FMCAD'07).

[5]  Scott Sanner,et al.  Symbolic Variable Elimination for Discrete and Continuous Graphical Models , 2012, AAAI.

[6]  Albert Oliveras,et al.  SMT Techniques for Fast Predicate Abstraction , 2006, CAV.

[7]  Raquel Urtasun,et al.  Efficient Inference of Continuous Markov Random Fields with Polynomial Potentials , 2014, NIPS.

[8]  Jesús A. De Loera,et al.  How to integrate a polynomial over a simplex , 2008, Math. Comput..

[9]  M. Tamer Özsu.  Synthesis Lectures on Data Management , 2010 .

[10]  Guy Van den Broeck,et al.  Component Caching in Hybrid Domains with Piecewise Polynomial Densities , 2016, AAAI.

[11]  J. Voelz.  Chapter 26 , 2019, The Crucible.

[12]  Prakash P. Shenoy,et al.  Inference in hybrid Bayesian networks using mixtures of polynomials , 2011, Int. J. Approx. Reason..

[13]  Toniann Pitassi,et al.  Combining Component Caching and Clause Learning for Effective Model Counting , 2004, SAT.

[14]  Steffen L. Lauritzen,et al.  Stable local computation with conditional Gaussian distributions , 2001, Stat. Comput..

[15]  Jesús A. De Loera,et al.  Software for exact integration of polynomials over polyhedra , 2011, ACCA.

[16]  Guy Van den Broeck,et al.  Probabilistic Inference in Hybrid Domains by Weighted Model Integration , 2015, IJCAI.

[17]  Adnan Darwiche,et al.  Compiling Probabilistic Graphical Models Using Sentential Decision Diagrams , 2013, ECSQARU.

[18]  Scott Sanner,et al.  Closed-Form Gibbs Sampling for Graphical Models with Algebraic Constraints , 2016, AAAI.

[19]  Alberto Griggio,et al.  The MathSAT 5 SMT Solver ⋆ , 2012 .

[20]  C. Caldwell.  Mathematics of Computation , 1999 .

[21]  Toniann Pitassi,et al.  Solving #SAT and Bayesian Inference with Backtracking Search , 2014, J. Artif. Intell. Res..