Skip to content

README.md

Latest commit

 

History

History
36 lines (27 loc) · 518 Bytes

README.md

File metadata and controls

36 lines (27 loc) · 518 Bytes

Kernel-PCA

A Python implementation of Kernel Principal Component Analysis (KPCA). Kernels implemented:

  • Linear
  • Radial Basis Function
  • Exponential
  • Laplacian
  • Anova
  • Polynomial
  • Sigmoid
  • Rotational quadratic
  • Multiquadric
  • Power
  • Spherical
  • Circular

Requirements

  • numpy
  • matplotlib
  • seaborn

Run

from kpca import KPCA
from kernels import kernel
X = np.array([[2,3,4], [1,2,3]]) # dxn
k = kernel(sigma=0.0009, d_anova=3).anova
kpca = KPCA(X, k, 3)
scores = kpca.project().T