Message-Passing Algorithms for Quadratic Programming Formulations of MAP Estimation

Computing maximum a posteriori (MAP) estimation in graphical models is an important inference problem with many applications. We present message-passing algorithms for quadratic programming (QP) formulations of MAP estimation for pairwise Markov random fields. In particular, we use the concave-convex procedure (CCCP) to obtain a locally optimal algorithm for the non-convex QP formulation. A similar technique is used to derive a globally convergent algorithm for the convex QP relaxation of MAP. We also show that a recently developed expectation-maximization (EM) algorithm for the QP formulation of MAP can be derived from the CCCP perspective. Experiments on synthetic and real-world problems confirm that our new approach is competitive with max-product and its variations. Compared with CPLEX, we achieve more than an order-of-magnitude speedup in solving optimally the convex QP relaxation.

[1]  D. Sontag.  1 Introduction to Dual Decomposition for Inference , 2010 .

[2]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[3]  Stephen P. Boyd,et al.  A tutorial on geometric programming , 2007, Optimization and Engineering.

[4]  Pradeep Ravikumar,et al.  Quadratic programming relaxations for metric labeling and Markov random field MAP estimation , 2006, ICML.

[5]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems , 1988 .

[6]  Michael I. Jordan,et al.  Graphical Models, Exponential Families, and Variational Inference , 2008, Found. Trends Mach. Learn..

[7]  Martin J. Wainwright,et al.  MAP estimation via agreement on (hyper)trees: Message-passing and linear programming , 2005, ArXiv.

[8]  Yair Weiss,et al.  Linear Programming Relaxations and Belief Propagation - An Empirical Study , 2006, J. Mach. Learn. Res..

[9]  M. .. Moore.  Exactly Solved Models in Statistical Mechanics , 1983 .

[10]  Tommi S. Jaakkola,et al.  Introduction to dual composition for inference , 2011 .

[11]  Shlomo Zilberstein,et al.  MAP Estimation for Graphical Models by Likelihood Maximization , 2010, NIPS.

[12]  Tommi S. Jaakkola,et al.  Tightening LP Relaxations for MAP using Message Passing , 2008, UAI.

[13]  Vladimir Kolmogorov,et al.  Analyzing Convex Relaxations for MAP Estimation , 2009 .

[14]  Vladimir Kolmogorov,et al.  Convergent Tree-Reweighted Message Passing for Energy Minimization , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.

[16]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1997 .

[17]  Alan L. Yuille,et al.  The Concave-Convex Procedure , 2003, Neural Computation.

[18]  Tommi S. Jaakkola,et al.  Fixing Max-Product: Convergent Message Passing Algorithms for MAP LP-Relaxations , 2007, NIPS.

[19]  Gert R. G. Lanckriet,et al.  On the Convergence of the Concave-Convex Procedure , 2009, NIPS.

[20]  Tommi S. Jaakkola,et al.  Learning Efficiently with Approximate Inference via Dual Losses , 2010, ICML.

[21]  Martin J. Wainwright,et al.  MAP estimation via agreement on trees: message-passing and linear programming , 2005, IEEE Transactions on Information Theory.