Industrial data asset label automatic classification method and system

The invention provides an industrial data asset label automatic classification method and system. The method comprises a classification model training link and a data classification application link. According to the method, an automatic label classification framework for continuous value characteristic attributes is designed on the basis of a probability density function in combination with a long-term memory vector, the problem of unbalanced sample category distribution is solved by adopting a hybrid sampling technology, and the automatic label classification effect of continuous value data assets under unbalanced samples commonly existing in the industry is achieved. The method can be widely applied to data multi-label automatic classification scenes influenced by multiple continuous value features in industrial scenes. 本发明提供了一种工业数据资产标签自动化分类方法和系统，包括：分类模型训练环节、数据分类应用环节。本发明基于概率密度函数，结合长期记忆向量设计了针对连续值特征属性的标签自动化分类架构，并采用混合采样技术解决样本类别分布不平衡的问题，实现工业中普遍存在的非均衡样本下的连续值数据资产的标签自动分类效果。本发明能够广泛应用于工业场景中针对受多个连续值特征影响的数据多标签自动化分类场景