Penentuan Kadar Gula Buah Salak Secara Tidak Merusak Menggunakan Near Infrared (Nir) Spectroscopy dan Interval Partial Least Squares (Ipls) Regression

Winarto Winarto (Politeknik Negeri Lampung)
Meinilwita Yulia (Politeknik Negeri Lampung)

TEKTAN
Vol 6 No 1 (2014)

The objectives of this research was to evaluate the feasibility of using NIR spectroscopy and interval partial least squares (iPLS) regression for nondestructive soluble solids content (SSC) determination in snake fruit. Spectra of snake fruit were acquired using a VIS-NIR USB4000 portable spectrometer in absorbance mode. The SSC was measured using a digital refractometer.  The calibration and validation models were developed using the iPLS regression using original spectra. The calibration results showed that the root mean square of cross-validation (RMSECV) of first six subintervals from 582-831 nm was lower than that of the average of RMSECV and these subintervals were selected for developing final iPLS regression. Using the selected subintervals the calibration has high quality of prediction with RPD = 3,35. Keywords: NIR spectroscopy, soluble solids content, absorbance mode, nondestructive   method, snake fruit, iPLS regression