作者: V. Centner , O.E. de Noord , D.L. Massart
DOI: 10.1016/S0003-2670(98)00543-1
关键词:
摘要: The present study categorises and compares several graphical numerical methods to detect the presence of nonlinearity in multivariate calibration. focus is on (quadratic) relationship between property interest (e.g. concentration) set instrumental absorbance) measurements. explored techniques are applied three simulated data sets where (non)linearity known, four experimental near infrared (NIR) sets. Mallows augmented partial residual plot recommended as most universal diagnostic nonlinearity. significance evaluated using runs test.