On the Interpolation between Product-Based Message Passing Heuristics for SAT
نویسنده
چکیده
This paper introduces a notational frame to characterize the four basic product-based Message Passing (MP) heuristics currently available for SAT: Belief Propagation (BP), Survey Propagation (SP), Expectation Maximization BP Global (EMBPG) and Expectation Maximization SP Global (EMSPG). Using this framework, the paper introduces indirect structural interpolation (ISI). Using this technique, we create a hierarchy of heuristics – each new level in this hierarchy consists of heuristics strictly more general than their predecessors. The final result is the ρσPMP heuristic, which is able to mimic all product-based MP heuristics and is hence a generalization for all them.
منابع مشابه
Applying Probabilistic Inference to Heuristic Search by Estimating Variable Bias University of Toronto Technical Report CSRG-577
Backbone variables have the same assignment in all solutions to a given constraint satisfaction problem; more generally, bias represents the proportion of solutions that assign a variable a particular value. Intuitively such constructs would seem important to efficient search, but their study to date has assumed a mostly conceptual perspective, in terms of indicating problem hardness or motivat...
متن کاملApplying Probabilistic Inference to Heuristic Search by Estimating Variable Bias
Backbone variables have the same assignment in all solutions to a given constraint satisfaction problem; more generally, bias represents the proportion of solutions that assign a variable a particular value. Intuitively such constructs would seem important to efficient search, but their study to date has assumed a mostly conceptual perspective, in terms of indicating problem hardness or motivat...
متن کاملSAT-Based Compositional Verification Using Lazy Learning
A recent approach to automated assume-guarantee reasoning (AGR) for concurrent systems relies on computing environment assumptions for components using the L algorithm for learning regular languages. While this approach has been investigated extensively for message passing systems, it still remains a challenge to scale the technique to large shared memory systems, mainly because the assumptions...
متن کاملProbabilistically Estimating Backbones and Variable Bias: Experimental Overview
Backbone variables have the same assignment in all solutions to a given constraint satisfaction problem; more generally, bias represents the proportion of solutions that assign a variable a particular value. Intuitively such constructs would seem important to efficient search, but their study to date has been from a mostly conceptual perspective, in terms of indicating problem hardness or motiv...
متن کاملSolving the Satisfiability Problem Using Message-passing Techniques
A Bayesian message-passing technique, the sum-product algorithm, is described and applied to the satisfiability problem, specifically 3-SAT. The results of this algorithm on randomly generated 3-SAT problems and standard SAT test sets are discussed and compared with conventional techniques. The algorithm is found to perform poorly in regions of the parameter space in which it is conventionally ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2013