Classification_LR.m

% Function to Cclassify data points using Logistic Regression classifier

function [Error,Predicted_Labels]=Classification_LR(Projected_Images_Training,Train_Labels,Projected_Images_Testing,Ground_Labels)

%Input
%
% Projected_Images_Training      : Matrix containing projected training Images 
% Projected_Images_Testing       : Matrix containing projected testing Images
% Train_Labels                   : Training Labesl of Images
% Ground_Labels                  : Ground Truth for each testing Image
%
% Output
% Error                          : Error of classification using Logistic Regression
% 
% 
% Author                         : Sunny Verma (sunnyverma.iitd@gmail.com)
% Last_Update                    : 28/08/2016
%

Testing_Images=size(Projected_Images_Testing,1);

t=templateLinear('Learner','logistic','Lambda','auto','Regularization','ridge','Solver','asgd');

LR_mdl=fitcecoc(Projected_Images_Training',Train_Labels,'Coding','onevsall','Learners',t,'ObservationsIn','columns');

Predicted_Labels=predict(LR_mdl,Projected_Images_Testing);

Error=nnz(Predicted_Labels-Ground_Labels)/Testing_Images;

end