TSUTA Mizuki, Principal Scientist, Nondestructive Evaluation Unit, Division of Analytical Science, Food Research Institute, NARO has received Best Award of Oral Presentation in The 7th Asian NIR Symposium. The awarding ceremony was held on February 13, 2020 at Thailand.
Step-wise Selectivity Ratio Method for Objective Variable Selection: A Case Study of Freshness Estimation for Leaf Vegetables
Japanese mustard spinach (Komatsuna) was stored at various temperatures and periods, and the visible-near infrared absorption spectra of the leaves was measured. A model for estimating the integrated temperature of the sample (product of storage temperature and number of days) from the absorption spectra was constructed by partial least squares (PLS) regression analysis. Integrated temperature is one of the indicators for freshness of vegetables. By selecting the wavelength conditions using the step-wise selectivity ratio method proposed by the authors, it was found that the estimation accuracy was improved compared to the case where all the wavelength conditions were used. Additionally, the absorption spectra at 1100-2500 nm of the leaf blade was found to be suitable for estimating the integrated temperature. Using published data on determination of octane number of gasoline using near infrared spectroscopy and determination of ovarian cancer by proteomics, it was shown that the step-wise selectivity ratio method is useful for improving estimation accuracy regardless of data.