Sebastian Raschka • 9/14/2014

Kernel tricks and nonlinear dimensionality reduction via RBF kernel PCA

This technical article introduces kernel methods for nonlinear dimensionality reduction, focusing on Gaussian Radial Basis Function (RBF) Kernel PCA. It contrasts linear PCA with kernel PCA, explains the kernel trick, and provides a step-by-step implementation guide. The content includes examples like half-moon shapes and the Swiss roll to demonstrate the technique's effectiveness on linearly inseparable data.

0 comments

#dimensionality reduction #Scikit Learn #Principal Component Analysis