Calibration transfer across near-infrared spectrometric instruments using Shenkʹs algorithm: effects of different standardisation samples

In this paper, transfer of calibration models between near-infrared spectrometric instruments using three different standardisation sets is described. The first set contains samples which are very similar to the agricultural samples from three different sets to be analysed, the second set contains generic standards, and the third one contains pure organic and inorganic chemicals. To test the accuracy of each standardisation, root mean square errors and correlation coefficients are computed before and after standardisation. To test the predictive ability, standard error of prediction for the three different prediction sets are also computed before and after standardisation. For standardisation, Shenkʹs algorithm is used: a description of this algorithm is given.