liangzh123

摘要： 左边的是线性可分的(Separable),margin = 2M = 2/||w||，右边的是线性不可分(Nonseparable)，其中标有 ξ 的点是在它们边界的另一边(未正确区分的点)，离边界 ξ∗j = Mξj 。所有正确可分的点有 ξ∗j = 0. 因此 ∑ ξ∗j 就是所有未正确区分点偏离距离，我们在最大化margin时要限制 ∑ ξ∗j ≤ constant C.我们用soft margin(a decision boundary that is tolerable to small training errors)，这个margin需要权衡margin的大小和允许错误区分点的个 阅读全文

摘要： Our training data consists of m tuples (x(1), y(1)), (x(2), y(2)), . . . , (x(m), y(m)), where x(i) = (x(i1), x(i2), …,x(id))T and y(i) ∈ {−1, 1},denote the class label. The hyperplane(decision boundary) of a linear classifier can be written in the following form: wTx + b = 0, where w and b are para 阅读全文