Random Features for Large-Scale Kernel Machines

← Back to document page

RD ω R2 x Kernel Name k(∆) p(ω) Gaussian e− k∆k2 2 2 (2π)− D 2 e− kωk2 2 2 Laplacian e−k∆k 1 Q d 1 π(1+ω2 d) Cauchy Q d 2 1+∆2 d e−k∆k 1 Figure 1: Random Fourier Features. Each component of the feature map z( x) projects onto a random

  Feature