Editing support vector machines
Haixin Ke, Xuegong Zhang · 2002
A support vector machine constructs an optimal hyperplane from a small set of samples near the boundary. This makes it sensitive to these specific samples and tends to result in machines either too complex with poor generalization ability or too imprecise with high training error, depending on the kernel parameters. In this paper, we present an improved version of the method, called editing support vector machine (ESVM), which removes some samples near the boundary from the training set. Experiments show that for cases that the two classes are overlapped, ESVM can get better generalizing ability, and ESVM is also more robust with noises.