The Disputed Federalist Papers : SVM Feature Selection via Concave Minimization. Glenn Fung and Olvi L. Mangasarian. CSNA 2002 June 13-16, 2002 Madison, Wisconsin. Outline of Talk. Support Vector Machines (SVM) Introduction. Standard Quadratic Programming Formulation.
Glenn Fung and Olvi L. Mangasarian
June 13-16, 2002
(Will concentrate on classification)
where e is a vector of ones.Algebra of the Classification Problem2-Category Linearly Separable Case
vectorsSupport Vector MachinesMaximizing the Margin between Bounding Planes
is the weight of the training error
s.t.Support Vector Machines: Linear Programming Formulation
normal to the separating hyperplane:
vector by the concave exponential:
Having , determine the next iterate
by solving the LP:
of steps (typically 5 to 7) at a stationary point.