Class RadialBaseKernel

  • All Implemented Interfaces:
    java.io.Serializable, SVMKernel

    public class RadialBaseKernel
    extends NestedKernel
    This kernel computes the radial base kernel that corresponds to a gausian distribution.

    The formula for this is exp( -||a - b|| / (2* width ^ 2)). The term a-b can be represented in an arbitrary feature space by using a nested kernel k, and becomes k(a, a) + k(b, b) - 2 * k(a, b).

    As k(x, x) is required repeatedly, I suggest using a DiagonalCachingKernel as the immediately nested kernel function.

    Author:
    Matthew Pocock
    See Also:
    Serialized Form
    • Method Detail

      • evaluate

        public double evaluate​(java.lang.Object a,
                               java.lang.Object b)
        Description copied from interface: SVMKernel
        Return the dot product of two vectors in an arbitrary feature space. In this implementation, the `vectors' can actually be arbitrary objects.
      • setWidth

        public void setWidth​(double width)
      • toString

        public java.lang.String toString()
        Overrides:
        toString in class java.lang.Object