基于中药提取的数据挖掘系统设计与实现

Design and Implementation of Data Mining System Based on Traditional Chinese Medicine Extracted

系统主要应用数据挖掘方法对中药提取数据进行分析和预测。首先对数据进行集成和离散化处理,得到适合数据挖掘的数据集,然后利用k-means和DBSCAN聚类算法对质检数据进行聚类,得到工艺参数质检区间;并对Apriori算法进行了改进,在算法中加入了用户兴趣度的概念,控制了候选集指数增长,得到工艺参数和固含量的关系;并利用三层BP神经网络算法训练网络模型,得出过程参数和结果质量参数的关系,发现数据中隐含的规律,为企业优化工艺以及提高其生产效率降低成本等提供科学的分析、决策辅助工具。

Data mining methods are applied to the system to extract data for analysis of Chinese medicine and forecasting.First,the data integration and discrete treatment are made,to obtain appropriate data mining data sets.Then k-means and DBSCAN clustering algorithm are used to cluster the data quality control.Quality control process parameters interval are obtained.Apriori algorithm is improved by adding a user interest degree in the concept.The candidate set of exponential growth is controlled.The relationship between parameters and solids is got.By using three-layer BP neural network algorithm trained network model,the relationship between the process parameters and results quality parameters is obtained.The law implied in the data is found.It provides a scientific analysis and decision support for enterprises to optimize processes and improve their productivity and reduce the costs.