Plot Support Vectors, Margin and decision boundary using LIBSVM MATLAB interface
4 views (last 30 days)
Show older comments
I am using Matlab-Libsvm Interface for binary classification using SVM. I have multi dimensional data. I need to plot decision boundary and margin along with support vectors.
Below is the code I used for this:
model = svmtrain(Y, X, sprintf('-c %f -w1 1 -w-1 5 -g %f', best_C, best_gamma));
w = model.SVs' * model.sv_coef;
b = -model.rho;
if (model.Label(1) == -1)
w = -w; b = -b;
end
y_hat = sign(w'*X' + b);
sv = full(model.SVs);
% plot support vectors
figure
plot(sv(:,1:14),'ko', 'MarkerSize', 10);
hold on
% plot decision boundary
plot_x = linspace(min(X(:,1)), max(X(:,1)), 30);
plot_y = (-1/w(2))*(w(1)*plot_x + b);
plot_z = sign(w(1)*plot_x + b)/norm(w);
plot(plot_x, plot_y, 'k-', 'LineWidth', 1)
hold on
plot(plot_x, plot_z, 'kx', 'LineWidth', 1);
hold off
hold off
But I am unable to plot decision boundary and margin using this code. Also, support vectors are plotted inappropriately.
Please let me know where I am wrong.
1 Comment
babita verma
on 19 Oct 2017
Hi Astha I am facing the same problem with plotting of the decision boundary. Were you able to resolve this issue?
regards
Answers (0)
See Also
Categories
Find more on Statistics and Machine Learning Toolbox in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!