NIRS Measurements. The NIRS analyse

NIRS Measurements. The NIRS analyses were carried out
on a dispersive near-infrared NIRSystems 5000 (Foss Instruments
Inc., Rodgau, Germany) in the reflectance mode. A
commercial spectral analysis program (NIRS2, Infrasoft International
Inc., Port Matilda, USA) was used to process the
data and to develop the most appropriate chemometric methods.
The spectral data were pretreated with weighted multiple
scatter correction (WMCS) to eliminate interferences of scatter
and were transformed individually with first- to third-derivative
processing. The calibration program was set up with a
wavelength range from 1108 to 2490 nm using the partial
least-squares (PLS) algorithm. The optimum number of PLS
factors used for the individual prediction was determined by
cross-validation (Martens and Naes, 1989). The accuracy of
the calibration models is described by the multiple coefficient
of determination (R2) and the overall error between modeled
and reference values [standard error of cross validation
(SECV)].

Results (Indonesian) 2:

NIRS Pengukuran. The NIRS analisis dilakukan
pada dispersif NIRSystems inframerah-dekat 5000 (Foss Instruments
Inc, Rodgau, Jerman) dalam modus reflektansi. Sebuah
program analisis spektral komersial (NIRS2, Infrasoft International
Inc, Port Matilda, USA) digunakan untuk memproses
data dan untuk mengembangkan metode chemometric yang paling tepat.
Data spektral yang pra-perawatan dengan tertimbang beberapa
koreksi pencar (WMCS) untuk menghilangkan gangguan dari menyebarkan
dan ditransformasikan secara individual dengan pertama-untuk ketiga derivatif
pengolahan. Program kalibrasi didirikan dengan
rentang panjang gelombang 1108-2490 nm menggunakan parsial
kuadrat-(PLS) algoritma. Jumlah optimum PLS
faktor digunakan untuk prediksi individu ditentukan oleh
cross-validasi (Martens dan Naes, 1989). Akurasi
model kalibrasi digambarkan oleh koefisien beberapa
determinasi (R2) dan kesalahan keseluruhan antara model
dan nilai-nilai referensi [standard error cross validasi
(SECV)].