A new perspective on principal component analysis using inverse covariance

Abstract EN

In his research paper we have proposed a new method of the orthogonal projection approximation for the extraction of he pri^pal eigenvectors of a data while using Z_1 inverse covariance regularization, thus naming the new method as Inverse Cov a ance Principal Component Analysis (ICPCA).

The basic idea lies in the mapping of an input space into a feature space v inverse covariance IT1 factorization and then computing the principal components in the extracted feature space.

The performa ce of the proposed method has been shown quantitatively and qualitatively on a well known the Essex University’s imag database.

The comparison shows that the proposed method outperforms competing eigenvalue decomposition (EVD) method (classical Principal Component Analysis) in variance coverage as well as in the execution time.