KEEPING-QUALITY ASSESSMENT OF PASTEURIZED MILK BY MULTIVARIATE-ANALYSIS OF DYNAMIC HEADSPACE GAS-CHROMATOGRAPHIC DATA .2. FLAVOR CLASSIFICATION BY LINEAR DISCRIMINANT-ANALYSIS
B. Vallejocordoba et S. Nakai, KEEPING-QUALITY ASSESSMENT OF PASTEURIZED MILK BY MULTIVARIATE-ANALYSIS OF DYNAMIC HEADSPACE GAS-CHROMATOGRAPHIC DATA .2. FLAVOR CLASSIFICATION BY LINEAR DISCRIMINANT-ANALYSIS, Journal of agricultural and food chemistry, 42(4), 1994, pp. 994-999
The flavor quality of pasteurized milk was classified by applying line
ar discriminant analysis (LDA) to volatiles as determined by dynamic h
easpace capillary gas chromatography (DH-GC). LDA was able to classify
milk into good, marginal, and poor quality or into fruity, rancid, an
d normal flavor groups at a greater than 80% success rate. The zone of
good or normal flavor groups could be visualized in two-dimensional c
anonical plots; thus, any sample deviating from this zone could be con
sidered a defective product. In addition, LDA of gas chromatographic d
ata allowed the association of milk volatiles that contributed most si
gnificantly to the discrimination among flavor quality groups. Spoilag
e-associated volatiles were identified by using DH-GC-MS. The analytic
al system established has potential for application in the dairy indus
try for early detection of spoilage or for tracing defects associated
with consumer complaints.