Journal of Aquatic Food Product Technology · 2012 · 45 citations · 23 references
Natural SciencesFood ScienceMarine BiologyIchthyologySpectroscopic AnalysisSpectroscopyInfrared Remote SensingInfrared SpectroscopyNear-infrared ImagingNear-infrared SpectroscopyOptical SpectroscopySpectroscopic AuthenticationBinary Logistic RegressionInfrared SpectraInfrared Spectroscopic AnalysisInfrared Spectroscopy AuthenticationLogistic RegressionNear-infrared Spectroscopy AssessmentPrincipal ComponentsVisible-near-infrared Spectroscopy
This study evaluated near-infrared (NIR) and visible-NIR (Vis-NIR) spectroscopy as a way to distinguish fresh (F) from frozen-thawed (T) swordfish cutlets (Xiphias gladius). A total of 90 F and 60 T samples were used. The T samples were stored at a high and low frozen temperature (HT: −10°C; LT: −18°C). Spectra were collected using a Vis-NIR portable spectrophotometer (380–1080 nm) and a NIR monochromator (1100–2500 nm). The percentage of correctly classified samples obtained with Vis-NIR spectroscopy was ≥ 96.7%, whereas that for NIR was ≥ 90.0%. The best classification was observed comparing F and HT samples using Vis-NIR (100 vs. 96.7%, respectively). The more descriptive principal component scores (PCS) of NIR and Vis-NIR were used with a multivariate binary logistic regression. The model with the PCS of the first two Vis-NIR principal components accounted for 81.1% of the classification. Vis-NIR could be a strategic tool to screen the cold treatment of swordfish.
23
Biostatistics 102: quantitative data--parametric & non-parametric tests.
Trade secrets: Renaming and mislabeling of seafood
Jennifer Jacquet, Daniel Pauly · Marine Policy · 2007 · 371 citations
Law, Fisheries, Food Science +12
Biostatistics 202: logistic regression analysis.