Near infrared quantitative analysis of total curcuminoids in rhizomes of Curcuma longa by moving window partial least squares regression

The present paper reports an application of moving window partial least squares regression (MWPLSR) to determine total curcuminoid content in rhizomes of Curcuma Longa by near infrared (NIR) spectroscopy. The MWPLSR method was applied to original and pretreated NIR data of Curcuma lenge rhizomes to select informative regions for total curcuminoids. Afterward, partial least squares (PLS) calibration models were developed and compared for each spectral region proposed by MWPLSR and the whole spectral region. The best PLS calibration model for total curcuminoids was obtained from 2nd derivative NIR spectra in the region of 2040-2486 nm. The standard error of prediction was 1.00% w/w and the ratio of prediction to deviation was 4.9 when using seven principle components. NIR spectroscopy combined with MWPLSR can lead to better calibration models with higher performance.