酿酒葡萄的等级判别方法研究

针对酿酒葡萄分类问题，在没有明确分类情况下，提出模糊C均值聚类和自主神经网络算法，在已知葡萄酒质量评分结果的前提下，首先对酿酒葡萄的理化指标和葡萄酒质量的数据进行极差归一化处理；运用模糊C均值聚算法得到分类隶属度矩阵和分类中心，并用自主神经网络无监督的学习方式成功地将酿酒葡萄分为三个等级．可以作为葡萄酒生产的参考。

The Wine Grape Level Discrimination Method

According to the problems of wine grape classification, in the case of no clear classifieation, this paper puts forward the fuzzy c- means clustering {FCM） and autonomous neural network algorithm, under ＂the circumstance of knowing wine quality score results ,first den with ＇the data of the physical and chemical index of wine grape and wine quality scores by normalization processing; By the fuzzy c-means clustering algorithm for classification can get membership degree matrix and classification center, under the condition of unsupervised learning way by the way of the autonomic nervous network wine grape is successfully divided into three levels. The ways can be used as a reference for grape wine production.