%Bayesian network based EDA for multi-objective 3-SAT %For details on the EDA application to the multi-objective 3-SAT problem, see (Santana_et_al:2009) global Formulas; % Global variable for SAT function load TypeForm_1_Form_1.mat; % Multi-objective SAT instance n = 20; % Number of variables m = 10; % Number of objectives c = 20; % Number of clauses PopSize = 500; NumbVar = n; cache = [1,1,1,1,1]; Card = 2*ones(1,NumbVar); maxgen = 50; F = 'EvaluateSAT'; % 3-SAT function selparams(1:2) = {0.5,'ParetoRank_ordering'}; sampling_params(1:3) = {PopSize,m,Obj_Card}; BN_params(1:7) = {'k2',5,0.05,'pearson','bayesian','no'}; edaparams{1} = {'learning_method','LearnBN',BN_params}; edaparams{2} = {'sampling_method','SampleBN',sampling_params}; edaparams{3} = {'selection_method','truncation_selection',selparams}; edaparams{4} = {'replacement_method','best_elitism',{'ParetoRank_ordering'}}; edaparams{5} = {'stop_cond_method','max_gen',{maxgen}}; [AllStat,Cache]=RunEDA(PopSize,NumbVar,F,Card,cache,edaparams) |