ccl_learna_sfa

 [model, stats] = ccl_learna_sfa (V, Un, model, search)

 Search the first dimension of constraint matrix alpha

 Input:
   V                                   UnUn(:)'
   Un                                  Observations of null space component
   model                               Model related parameters
   search                              Searching configurations

 Output:
   model                               Returned model
   stats                               Mean

0001 function [model, stats] = ccl_learna_sfa (V, Un, model, search)
0002 % [model, stats] = ccl_learna_sfa (V, Un, model, search)
0003 %
0004 % Search the first dimension of constraint matrix alpha
0005 %
0006 % Input:
0007 %   V                                   UnUn(:)'
0008 %   Un                                  Observations of null space component
0009 %   model                               Model related parameters
0010 %   search                              Searching configurations
0011 %
0012 % Output:
0013 %   model                               Returned model
0014 %   stats                               Mean
0015 
0016 for i = 1:search.dim_s
0017     alpha         = search.alpha{i} ;
0018     AA            = pinv(alpha)*alpha ;
0019     stats.umse(i) = sum ( V*AA(:) ) ;
0020 end
0021 [min_err, min_ind]  = min(stats.umse) ;
0022 model.theta         = search.theta{min_ind} ;
0023 model.alpha         = search.alpha{min_ind} ;
0024 model.P             = search.I_u - pinv(model.alpha) * model.alpha ;
0025 model.variance      = sum(var( model.P*Un, 0, 2)) ;
0026 model.umse_j        = stats.umse(min_ind) / search.dim_n ;
0027 model.nmse_j        = model.umse_j        / model.variance ;
0028 end