Predicting Milk Shelf‐life Based on Artificial Neural Networks and Headspace Gas Chromatographic Data
Predicting Milk Shelf‐life Based on Artificial Neural Networks and Headspace Gas Chromatographic Data
复制标题
DOI：
10.1111/j.1365-2621.1995.tb06253.x
复制
发表时间：
1995-09
影响因子：
3.9
通讯作者：
B. Vallejo‐Cordoba;G. Arteaga;S. Nakai
中科院分区：
文献类型：
--
作者：
B. Vallejo‐Cordoba;G. Arteaga;S. Nakai
人工神经网络（ANN）的有用性进行了评估，并比较主成分回归（PCR）的牛奶货架期预测的气相色谱配置文件和风味相关的货架期的多变量解释。训练集包括巴氏杀菌奶储存期间收集的动态顶空气相色谱数据（用于做出决定的神经网络的输入信息）及其相应的保质期（预测或响应）。人工神经网络的预测性优于PCR。由实验值与预测值的回归分析得出的货架期2天的估计值的标准误差表明ANN具有较高的可预测性。
The usefulness of artificial neural networks (ANN) for milk shelf-life prediction by multivariate interpretation of gas chromatographic profiles and flavor-related shelf-life was evaluated and compared to principal components regression (PCR). The training set consisted of dynamic headspace gas chromatographic data collected during storage of pasteurized milk (input information for the neural network used to make a decision) and its corresponding shelf-life (prediction or response). ANN had better predictability than PCR. A standard error of the estimate of 2 days in shelf-life resulting from regression analysis of experimental vs predicted values indicated a high predictability of ANN.