Journal
ANALYTICAL LETTERS
Volume 54, Issue 14, Pages 2304-2314Publisher
TAYLOR & FRANCIS INC
DOI: 10.1080/00032719.2020.1858092
Keywords
miso; partial least-square regression (PLSR; quality classification; sensory evaluation; Visible; near-infrared (Vis; NIR) spectroscopy)
Categories
Ask authors/readers for more resources
This study developed a rapid method for predicting the sensory qualities of miso products using visible/near-infrared spectroscopy combined with a partial least-square regression algorithm, demonstrating its effectiveness and feasibility. The best performing model utilized the first derivative pretreatment of the spectra, effectively classifying miso products and potentially serving as a low-cost and nondestructive quality assessment tool.
Miso, the Japanese fermented soybean paste, an essential seasoning for preparing washoku, is gaining increased international recognition. Miso quality is normally evaluated by a panel of sensory assessors, which is a time-consuming and costly procedure. The final quality of miso products depends on the complex interactions of color, taste, flavor, and aroma. This study aimed to use visible/near-infrared (Vis/NIR) spectroscopy combined with a partial least-square regression (PLSR) algorithm to develop a rapid method for predicting the sensory qualities of miso products. The full Vis/NIR spectrum with nine different pretreatment methods was used to build calibration, cross validation and prediction models. The best performance was achieved by a PLSR model with the first derivative pretreatment of the spectra, giving the root mean square error of prediction (RMSEP) and bias values of 25 and 13, respectively, for the validation test. This model effectively classified miso products into different grades. An investigation of particular spectral regions revealed that a similar performance to the best PLSR model was obtained using just the spectra from 400 to 1100 nm with first derivative pretreatment or combined with multiplicative scatter correction (MSC), and standard normal variate (SNV) methods. The study demonstrated that Vis/NIR spectroscopy has the potential for evaluating miso quality with the advantages of being rapid, accurate, low cost, and nondestructive.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available