Abstract
This paper introduces the concept of <i>eigenvalue manipulating transformation</i> (EMT) of a data matrix for noise suppression in two-dimensional (2D) correlation spectroscopy. The FT-IR spectra of a polystyrene/methyl ethyl ketone/toluene solution mixture during the solvent evaporation process, to which were added a substantial amount of artificial noise, have been analyzed. By uniformly raising the power of a set of eigenvalues, the major eigenvalues become more prominent. As a consequence, minor eigenvectors representing the noise component are no longer strongly represented in the reconstructed data. This EMT operation is similar to the simple truncation of noise-dominated minor factors practiced in standard principal component analysis (PCA), as demonstrated in our preceding paper on PCA-2D correlation spectroscopy. The effect of this new EMT scheme is more gradual, with attractive flexibility to continuously fine-tune the balance between the desired noise reduction effect and the retention of pertinent spectral information.
PDF Article
More Like This
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription