An electronic nose based on ultra-fast gas chromatography (UFGC E-nose) was used to analyze the volatile flavor components of 24 samples of whole milk powder for infants, students, adults and middle-aged and elderly people. The aroma characteristics of milk powders for different age groups were classified and predicted by linear discriminant analysis (LDA) and linear support vector machine (LSVM). The results showed that a total of 54 volatile flavor components were detected in all samples, among which 37 components were common to these samples including propionaldehyde, acetone, hexanal, and caproic acid. Methyl propionate and 1-octen-3-one were exclusively found in infant milk powder. (E,E)-2,4-Decadienal and γ-unsecalactone were unique to students’ milk powder. Heptanal was detected only in adult milk powder. 2-Butanone and benzaldehyde were only found in middle-aged and elderly milk powder. In total, 23 volatile components contributing to the aroma were identified according to their relative odor activity values and aroma radar map, and 10 of them were the key aroma compounds in milk powders for the four age groups, all of which had milky, vegetable-like, fruity, flowery, grassy, caramel-like and tar-like aromas. In addition, infant milk powder had a characteristic apricot kernel-like aroma. The sensory evaluation results showed that the flavor of students’ milk powder and adult milk powder was better than that of infant milk powder and middle-aged and elderly milk powder, which was significantly different from the aroma radar map results. The classification and prediction accuracy of LDA and LSVM for milk powder for different age groups were 93.3% and 94.2%, respectively.

An ultra-fast gas chromatography-based electronic nose was used to qualitatively and quantitatively analyze the volatile flavor components in 36 samples of whole milk, low-fat milk, and skimmed milk subjected to ultra-high temperature (UHT) treatment. Using a combination of principal component analysis (PCA) and multi-layer perceptron neural network (MLP-NN), UHT-treated milk samples were classified and predicted according to their aroma characteristics. The results showed that 29 volatile flavor components were detected in all milk samples, mainly including acetone, n-butanol, and δ-decalactone, acetone and decanal being the most and least abundant, respectively. Isobutanol, acetoin, 1-pentanol, E-3-hexenal and decanal only existed in UHT-treated whole milk. 3-Methylheptane, 2,6-dimethylpyrazine, and E-2-nonen-1-ol were exclusively detected in UHT-treated low-fat milk. α-Pinene, 5-methylfurfural and decanoic acid were only detected in UHT-treated skimmed milk. The major aroma characteristics of UHT-treated milk were vegetable-like, creamy, grassy, fruity, and malty aroma notes. Among them, vegetable-like, creamy and malty aromas were the most widely distributed in UHT-treated whole fat milk. Fruity aroma was the most widely distributed in UHT-treated low-fat milk. Grassy aroma was the most widely distributed in UHT-treated skimmed milk. PCA showed that there was a significant difference between different brands of UHT-treated milk. The prediction accuracy of MLP-NN for UHT-treated milk was as high as 98.6%.