ccl_error_nmse

 [nmse v mse] = ccl_error_nmse(Y,Yp)
 
 Calculate normalised mean square error

 Input:
   
   Y                      Target data points
   Yp                     Predictions

 Output:
 
   nmse                   Normalised mean square error
   v                      Variance in the observations
   mse                    Mean square error

0001 function [nmse v mse] = ccl_error_nmse(Y,Yp)
0002 % [nmse v mse] = ccl_error_nmse(Y,Yp)
0003 %
0004 % Calculate normalised mean square error
0005 %
0006 % Input:
0007 %
0008 %   Y                      Target data points
0009 %   Yp                     Predictions
0010 %
0011 % Output:
0012 %
0013 %   nmse                   Normalised mean square error
0014 %   v                      Variance in the observations
0015 %   mse                    Mean square error
0016 
0017 % CCL: A MATLAB library for Constraint Consistent Learning
0018 % Copyright (C) 2007  Matthew Howard
0019 % Contact: matthew.j.howard@kcl.ac.uk
0020 %
0021 % This library is free software; you can redistribute it and/or
0022 % modify it under the terms of the GNU Lesser General Public
0023 % License as published by the Free Software Foundation; either
0024 % version 2.1 of the License, or (at your option) any later version.
0025 %
0026 % This library is distributed in the hope that it will be useful,
0027 % but WITHOUT ANY WARRANTY; without even the implied warranty of
0028 % MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
0029 % Lesser General Public License for more details.
0030 %
0031 % You should have received a copy of the GNU Library General Public
0032 % License along with this library; if not, write to the Free
0033 % Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
0034 
0035 N    = size(Y,2);          % get no. data points
0036 mse  = sum((Y-Yp).^2,2)/N; % get mean squared error
0037 v    = var(Y,0,2);         % get variance
0038 nmse = mse/v;              % compute nmse
0039