Source code for vl_demo_svm.m

This file is located in the toolbox/demo folder in VLFeat package.

% VL_DEMO_SVM  Demo: SVM: 2D linear learning
function vl_demo_svm
y=[];X=[];

% Load training data X and their labels y
load('vl_demo_svm_data.mat')


Xp = X(:,y==1);
Xn = X(:,y==-1);

figure
plot(Xn(1,:),Xn(2,:),'*r')
hold on
plot(Xp(1,:),Xp(2,:),'*b')
axis equal ;
vl_demo_print('svm_training') ;
% Parameters
lambda = 0.01 ; % Regularization parameter
maxIter = 1000 ; % Maximum number of iterations

energy = [] ;
% Diagnostic function
function diagnostics(svm)
  energy = [energy [svm.objective ; svm.dualObjective ; svm.dualityGap ] ] ;
end

% Training the SVM
energy = [] ;
[w b info] = vl_svmtrain(X, y, lambda,...
                           'MaxNumIterations',maxIter,...
                           'DiagnosticFunction',@diagnostics,...
                           'DiagnosticFrequency',1)

% Visualisation
eq = [num2str(w(1)) '*x+' num2str(w(2)) '*y+' num2str(b)];

line = ezplot(eq, [-0.9 0.9 -0.9 0.9]);
set(line, 'Color', [0 0.8 0],'linewidth', 2);

vl_demo_print('svm_training_result') ;


figure
hold on
plot(energy(1,:),'--b') ;
plot(energy(2,:),'-.g') ;
plot(energy(3,:),'r') ;
legend('Primal objective','Dual objective','Duality gap')
xlabel('Diagnostics iteration')
ylabel('Energy')
vl_demo_print('svm_energy') ;

end