采购决策支持系统中模型自动选择技术的研究  

Technology Research of model automatic selection in Decision Support System

作　　者：邵丽丽[1] 

机构地区：[1]菏泽学院计算机与信息工程系,山东菏泽274015

出　　处：《微计算机信息》2012年第6期145-146,175,共3页Control & Automation

摘　　要：针对采购者往往不是计算机专业人员,不知道如何根据药品销售历史数据的特点在采购决策支持系统中选取适当的模型来进行预测的问题,本文采用一种基于遗传算法求解树结点的线性权值矢量,并根据树节点的错误率与分割后的错误率减少量构造二元决策树的方法,来实现采购决策支持系统中模型自动选择,最后仿真实验证明文中方法构造的二元决策树能正确的完成模型的自动选择。Buyers are often not computer professionals, they don＇t; know how to select proper model to forecast according to drug sales history dates. This paper uses genetic algorithm to solve tree node＇s linear weights vector, to construct binary decision tree based on tree node＇s error rate and the error rate decrease after division, to realize model automatic selection in purchasing Decision Support System. The simulation results have proven the constructed binary decision tree can correctly realize model automatic selection.

关 键 词：药品采购决策支持系统 遗传算法 二元决策树 模型 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]