Training of support vector machine (SVM) requires the solution of a very large quadratic programming (QP) problem. This paper proposes an algorithm for training SVM's: Sequential Minimal Optimization, or SMO. SMO breaks the large QP problem into a series of smallest possible QP problems which are analytically solvable. Thus, SMO does not require a numerical QP library. SMO's computation time is dominated by evaluation of the kernel, hence kernel optimizations substantially quicken SMO. For the MNIST database, SMO is 1.7 times as fast as PCG chunking; while for the UCI adult databse and linear SVM's, SMO can 1500 times faster than the PCG chunking algorithm.
In Proc. Advances in Neural Information Processing Systems 11
John Platt. Sequential Minimal Optimization: A Fast Algorithm for Training Support Vector Machines, April 1998.
John C. Platt. Fast Training of Support Vector Machines Using Sequential Minimal Optimization, MIT Press, January 1998.