Evaluation of the mineral contents in fish meal by FT- NIR using PLS and Kernel PLS

In the present work we study the use of Fourier transform near infrared spectroscopy (FTNIRS) technique to analysis the calcium (Ca), phosphorus (P) and copper (Cu) contents of fish meal. The regression methods employed were partial least squares (PLS) and kernel partial least squares (KPLS). The results showed that the efficiency of KPLS was better than PLS. As a whole, the application of FT-NIRS with PLS and KPLS was as a suitable option for replacing the routine chemical analysis to assess the mineral content in fish meal, allowing immediate control of the fish meal without prior sample treatment or destruction.