Training Support Vector Machines using Gilbert's Algorithm Shawn Martin   View Paper (PDF)   Return to Session 13: Learning Techniques II Abstract Support Vector Machines are classifiers designed around the computation of an optimal separating hyperplane. This hyperplane is typically obtained by solving a constrained quadratic programming problem, but may also be located by solving a nearest point problem. Gilbertˇ¦s Algorithm can be used to solve this nearest point problem but is unreasonably slow. In this paper we present a modified version of Gilbertˇ¦s Algorithm for the fast computation of the Support Vector Machine hyperplane. We then compare our algorithm with the Nearest Point Algorithm and with Sequential Minimal Optimization. ©2006 Association for Computing Machinery