brede_nmf_cost_eucl

 brede_nmf_cost_eucl  - NMF euclidean cost

       function cost = brede_nmf_cost_eucl(A,B);
       function cost = brede_nmf_cost_eucl(A,C,D);

       Input:  A     Matrix (target)
               B     Matrix (model output)
               C     Factorized matrix
               D     Factorized matrix

       Output: cost  Costfunction value

       The function computes the euclidean costfunction in connection
       with non-negative matrix factorization.

                  cost = ||A-B||^2

       or with A, C and D input arguments:

                  cost = || A - C*D ||^2

       NaN elements are ignored.

       If the function is called with 3 arguments and the A matrix is
       sparse the the cost is computed iteratively over columns of D
       to avoid memory consumption problems.

       Ref: DD Lee, HS Seung, Algorithms for Non-negative Matrix
       Factorization, 2000, Equation 2.

       See also BREDE, BREDE_MAT, BREDE_NMF_COST_DIV, BREDE_NMF_EST.

 $Id: brede_nmf_cost_eucl.m,v 1.4 2006/04/24 15:22:16 fn Exp $