| In their basic form, SVMs attempt to perform classification by constructing hyperplanes in a multidimensional space that separates the cases of different class labels. |
| In Section 4 we show that any PMD in which a coplane is not the intersection of two hyperplanes is an extension of a projective plane. |
| The heart of this paper is the following correspondence between Shi and Ish hyperplanes. |
| Its goal is to find hyperplanes in the attribute's space in order to maximize the margin between instances that belong to distinct classes. |
| It then employs hyperplanes to separate positive data from the negative ones. |
