目的探究白酒感官品评与白酒风味成分之间的关系,实现通过感官品评对风味成分进行预测。方法采用变分自编码器(variational auto encoder,VAE)对原始数据进行增强,以多核支持向量回归(multi-kernel support vector regression,MKSVR)结...目的探究白酒感官品评与白酒风味成分之间的关系,实现通过感官品评对风味成分进行预测。方法采用变分自编码器(variational auto encoder,VAE)对原始数据进行增强,以多核支持向量回归(multi-kernel support vector regression,MKSVR)结合遗传算法(genetic algorithm,GA)建立单预测模型,再采取逐步预测的方式按照酸、酯、醇、醛类物质的顺序进行预测,从而构建最终模型。结果在经过VAE对数据进行增强的条件下,多元线性回归(mixed logistic regression,MLR)对酸、酯、醇、醛类物质预测的拟合优度分别为0.9660、0.9106、0.8767、0.8686,随机森林(random forests,RF)对酸、酯、醇、醛类物质预测的拟合优度分别为0.9663、0.9186、0.8805、0.8708,GA-MKSVR对酸、酯、醇、醛类物质预测的拟合优度分别为0.9715、0.9423、0.9072、0.8809,GA-MKSVR逐步预测对酸、酯、醇、醛类物质预测的拟合优度分别为0.9715、0.9447、0.9102、0.8851,GA-MKSVR逐步预测的效果均为最优。结论GA-MKSVR逐步预测方法相较于传统的机器学习方法,具有更好的性能,对数据具有更高的适应性,能更好地构建白酒感官与风味成分之间的关系模型。展开更多
Green tea is a popular product with a high consumption in Vietnam. Moreover, green tea produced in Tan Cuong commune, Thai Nguyen province, has long been recognized for its superior quality to those coming from other ...Green tea is a popular product with a high consumption in Vietnam. Moreover, green tea produced in Tan Cuong commune, Thai Nguyen province, has long been recognized for its superior quality to those coming from other regions of the country. This study aims at comparing the tea from Thai Nguyen and Phu Tho and finding out if the difference can be identified through sensory analysis. Two products picked from Tan Cuong, Thai Nguyen province, and two others from Phu Ho district, Phu Tho province were evaluated by a trained panel of 12 judges （11 women and 1 man）. The sensory evaluation of the dry tea （11 descriptors） was carried out separately from the brewed tea （21 descriptors） and brewed leaf （five descriptors） using an unstructured intensity scale （10 cm）. Statistic analyses have shown differences between regions in all three groups of attributes： dry leaf, liquor and brewed leaf.展开更多
The impact of different phases of shipment （at sea and at port） on two German white wines of two vintages and the lasting effects of the temperature regimes over time was investigated. The wines were subjected to th...The impact of different phases of shipment （at sea and at port） on two German white wines of two vintages and the lasting effects of the temperature regimes over time was investigated. The wines were subjected to three temperature programs--control （15 ℃）, linear increase （15℃ steadily increasing to 45 ℃）, and diurnal fluctuation （15 ℃/40 ℃）--in both movement and non-movement conditions. The wines were analyzed for chemical, physical and sensorial changes at one and eight months post-treatment. Changes in temperature and pressure were recorded within the bottles, which correlated with the temperature programs： ＋0.04 bar/℃ in the linear increase program and ＋0.08 bar/℃ in the diurnal fluctuation program. The oxygen levels in the headspace and in the wine were monitored during all of the treatments. The oxygen development in the bottles was similar between the diurnal and linear programs, and was found to be distinctive from the control program. The chemical analysis revealed that there were significant differences related to the experimental treatments of the wines for the following parameters： tartaric acid, free sulfur dioxide, total sulfur dioxide and percent cork weight loss measurements. Difference sensory testing found very few differences. After eight months storage, significant differences were found in the Diurnal Non-movement treatment compared to Linear Non-movement and control treatments, as well as Diurnal Movement and Control treatments for the 2014 Miiller-Thurgau wine. Sensory descriptive analysis of the wines found that the wines could be differentiated by variety, but could not be distinguished according to experimental treatment after one month storage. These results indicate that wines of these types are more robust to shipping conditions than previously found.展开更多
文摘目的探究白酒感官品评与白酒风味成分之间的关系,实现通过感官品评对风味成分进行预测。方法采用变分自编码器(variational auto encoder,VAE)对原始数据进行增强,以多核支持向量回归(multi-kernel support vector regression,MKSVR)结合遗传算法(genetic algorithm,GA)建立单预测模型,再采取逐步预测的方式按照酸、酯、醇、醛类物质的顺序进行预测,从而构建最终模型。结果在经过VAE对数据进行增强的条件下,多元线性回归(mixed logistic regression,MLR)对酸、酯、醇、醛类物质预测的拟合优度分别为0.9660、0.9106、0.8767、0.8686,随机森林(random forests,RF)对酸、酯、醇、醛类物质预测的拟合优度分别为0.9663、0.9186、0.8805、0.8708,GA-MKSVR对酸、酯、醇、醛类物质预测的拟合优度分别为0.9715、0.9423、0.9072、0.8809,GA-MKSVR逐步预测对酸、酯、醇、醛类物质预测的拟合优度分别为0.9715、0.9447、0.9102、0.8851,GA-MKSVR逐步预测的效果均为最优。结论GA-MKSVR逐步预测方法相较于传统的机器学习方法,具有更好的性能,对数据具有更高的适应性,能更好地构建白酒感官与风味成分之间的关系模型。
文摘Green tea is a popular product with a high consumption in Vietnam. Moreover, green tea produced in Tan Cuong commune, Thai Nguyen province, has long been recognized for its superior quality to those coming from other regions of the country. This study aims at comparing the tea from Thai Nguyen and Phu Tho and finding out if the difference can be identified through sensory analysis. Two products picked from Tan Cuong, Thai Nguyen province, and two others from Phu Ho district, Phu Tho province were evaluated by a trained panel of 12 judges （11 women and 1 man）. The sensory evaluation of the dry tea （11 descriptors） was carried out separately from the brewed tea （21 descriptors） and brewed leaf （five descriptors） using an unstructured intensity scale （10 cm）. Statistic analyses have shown differences between regions in all three groups of attributes： dry leaf, liquor and brewed leaf.
文摘The impact of different phases of shipment （at sea and at port） on two German white wines of two vintages and the lasting effects of the temperature regimes over time was investigated. The wines were subjected to three temperature programs--control （15 ℃）, linear increase （15℃ steadily increasing to 45 ℃）, and diurnal fluctuation （15 ℃/40 ℃）--in both movement and non-movement conditions. The wines were analyzed for chemical, physical and sensorial changes at one and eight months post-treatment. Changes in temperature and pressure were recorded within the bottles, which correlated with the temperature programs： ＋0.04 bar/℃ in the linear increase program and ＋0.08 bar/℃ in the diurnal fluctuation program. The oxygen levels in the headspace and in the wine were monitored during all of the treatments. The oxygen development in the bottles was similar between the diurnal and linear programs, and was found to be distinctive from the control program. The chemical analysis revealed that there were significant differences related to the experimental treatments of the wines for the following parameters： tartaric acid, free sulfur dioxide, total sulfur dioxide and percent cork weight loss measurements. Difference sensory testing found very few differences. After eight months storage, significant differences were found in the Diurnal Non-movement treatment compared to Linear Non-movement and control treatments, as well as Diurnal Movement and Control treatments for the 2014 Miiller-Thurgau wine. Sensory descriptive analysis of the wines found that the wines could be differentiated by variety, but could not be distinguished according to experimental treatment after one month storage. These results indicate that wines of these types are more robust to shipping conditions than previously found.
