Delaunay triangulation method for multivariate calibration

Abstract

Listen

Translate article

The Delaunay triangulation (DT) method is proposed as a new local multivariate calibration method. DT was developed within computational geometry, and it is shown that it has potential for applications in analytical chemistry, such as multivariate calibration. The study compares the performance of the DT method with the global methods principal component regression (PCR) and partial least squares (PLS) and the local methods locally weighted regression (LWR) and the law of mixtures (LM) method. For the datasets studied the DT method gives similar results when the root mean square error for prediction (RMSEP) value is compared. However, the DT method requires fewer components than PCR and PLS.