= Support-vector leave-one-out error bound
{title2=$\operatorname{err}_{CV}\leq s/n$}
For a fixed $C$ in the unnormalized sum-of-slacks objective and unique training optimizers, deleting an observation with $\alpha_i=0$ preserves the <Karush-Kuhn-Tucker conditions> and the fitted decision function. It is correctly classified when held out. Consequently <Leave-one-out cross-validation> makes at most $s$ errors, where $s$ counts positive dual coefficients. With geometric support vectors, counting all points of signed margin at most one also gives the bound, without a nondegeneracy assumption.
Back to article page
