MCSAT Class Reference

MC-SAT is an MCMC inference algorithm designed to deal efficiently with probabilistic and deterministic dependencies (See Poon and Domingos, 2006). More...

#include <mcsat.h>

Inheritance diagram for MCSAT:

MCMC Inference List of all members.

Public Member Functions

 MCSAT (VariableState *state, long int seed, const bool &trackClauseTrueCnts, MCSatParams *mcsatParams)
 Constructor: Constructs unit propagation and SampleSat.
 ~MCSAT ()
 Destructor: Instances of unit propagation and MaxWalksat are deleted.
void init ()
 Initializes MC-SAT with MaxWalksat on hard clauses.
void infer ()
 Performs MC-SAT inference.

Detailed Description

MC-SAT is an MCMC inference algorithm designed to deal efficiently with probabilistic and deterministic dependencies (See Poon and Domingos, 2006).

It wraps a procedure around SampleSat, thus enabling it to sample nearly uniform.

Definition at line 82 of file mcsat.h.


The documentation for this class was generated from the following file:
Generated on Tue Jan 16 05:30:06 2007 for Alchemy by  doxygen 1.5.1