Ensemble Learning Based Multiple Kernel Principal Component Analysis for Dimensionality Reduction and Classification of Hyperspectral Imagery

Algorithm 2

M-KPCA Algorithm.

Input: Given training set with labels .

(a) Obtain ’s for corresponding pre-selected kernels using Eqs. (12) and (13).

(b) Create the kernel matrix for each as in Eq. (9).

(c) Calculate the eigenvectors and eigenvalues of . Sort the eigenvectors according to

the decreasing order of and select first eigenvectors .

(d) Calculate the eigenvectors and eigenvalues of . Sort the eigenvectors according to

the decreasing order of and select first eigenvectors .

(e) The final subspaces are and .

Output: and .