Использование метода PLS-дискриминаций для классификации исходного и модифицированного торфа

Abstract

The use of partial least squares projections to latent structures with discriminant analysis (PLSDA) in classification of initial and modified peat by a set of properties is presented. Modification of peat is achieved by heating. The set of peat properties such as component composition, concentration of paramagnetic centers and IR-spectra was used. It is shown that a developed PLS-DA classification model allows estimating peat class (initial or modified) by a set of properties without prediction errors by using two or three latent variables.