How " label = predict(SVMModel,X) " computes the labels for 'rbf' kernel?

3 views (last 30 days)
Hello, Can anyone please direct me to some documentation regarding the method of labels-computation for 'rbf' kernel using "predict" function? I assume it does not just do f(x) = sum(alpha(i)*y(i)*rbf(x-x(i))) + bias, for all support-vectors x(i), because I wrote my own version which does just that (using the same scale, same alpha and everything else, which I got by running "fitcsvm"), but got different labels! Your help will be highly appreciated, Nadav

Answers (0)

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!