EXAMPLE 7: Gaussian UMDA for the Offline HP Model continuous function For reference on the Offline HP model see: -- H. P. Hsu, V. Mehra and P. Grassberger (2003) Structure optimization in an off-lattice protein model. -- Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Sep;68(3 Pt 2):037703. Epub 2003 Sep 30. -- http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PLEEE8000068000003037703000001&idtype=cvips&gifs=yes
0001 % EXAMPLE 7: Gaussian UMDA for the Offline HP Model continuous function 0002 % For reference on the Offline HP model see: 0003 %-- H. P. Hsu, V. Mehra and P. Grassberger (2003) Structure optimization in an off-lattice protein model. 0004 %-- Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Sep;68(3 Pt 2):037703. Epub 2003 Sep 30. 0005 %-- http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PLEEE8000068000003037703000001&idtype=cvips&gifs=yes 0006 0007 Fibbonacci_n = 7; % Fibbonacci_n: Value n for the construction of the Fibbonacci sequence. NumbVar = F(n) 0008 global HPInitConf; 0009 HPInitConf = CreateFibbInitConf(Fibbonacci_n); % HP Fibbonacci configuration 0010 NumbVar = size(HPInitConf,2); 0011 PopSize = 1000; 0012 F = 'EvaluateOffHPProtein'; 0013 cache = [0,0,0,0,0]; Card = [zeros(1,NumbVar);2*pi*ones(1,NumbVar)]; 0014 edaparams{1} = {'learning_method','LearnGaussianUnivModel',{}}; 0015 edaparams{2} = {'sampling_method','SampleGaussianUnivModel',{PopSize,3}}; 0016 edaparams{3} = {'replacement_method','elitism',{1,'fitness_ordering'}}; 0017 edaparams{4} = {'selection_method','prop_selection',{2}}; 0018 edaparams{5} = {'repairing_method','SetInBounds_repairing',{}}; 0019 [AllStat,Cache]=RunEDA(PopSize,NumbVar,F,Card,cache,edaparams) 0020 % To draw the resulting solution use function OffPrintProtein(vector), 0021 % where vector is the best solution found.