基于广义粗集和知识共享的供应链需求预测R5模型

在知识共享前提下,提出了供应链需求预测的案例推理R5模型.此模型以粗集方法进行知识发现,建造和划分案例库,以案例推理管理和应用知识,将广义粗集的数据挖掘功能引入案例推理模式当中,用于指导供应链企业从过去的合作经验中有效学习其合作伙伴的“预测知识”,进而利用新的需求信息进行当前产品的需求预测模型选择.仿真算例表明,将变精度粗集和广义相似关系这两种粗集的泛化形式引入案例推理机制,可以提高推理效率和容错性;利用数据挖掘技术,从相关产品的环境信息及供应链成员的特征参数中发现需求预报的领域知识与深层知识,可以降低知识获取成本及模型选择的主观性.

R5 Model to Share Demand Forecast in a Supply Chain Based on Generalized Rough Set and Knowledge Sharing

This paper puts forward a R~5 model for sharing the knowledge of demand forecasts in a supply chain, which discovers knowledge, builds and partitions case base with rough set means, manages and applies knowledge with case based reasoning (CBR), so as to introduce the data mining function of rough set to the pattern of CBR. By using this model, the supply chain number can make the best use of informauion on hand to learn his cooperators' forecast knowledge, and to choose a forecast model for a coming product. Furthermore, an example is given to indicate that the model can improve the reasoning efficiency and reduce the cost and subjectivity of model choice.