Nonlinear Kernel-Based Chemometric Tools: a Machine Learning Approach

Abstract

This paper provides a short introduction to support vector machines and other nonlinear kernel-based methods recently developed in machine learning research. We describe principles of construction of the nonlinear kernel-based variants of linear methods, which have been widely used in the domain of chemometrics. These include nonlinear kernel forms of the partial least squares, canonical correlation analysis, principal component analysis, principal component regression and ridge regression methods.


Go back