摘要
决策树方法一直被用于贝叶斯决策问题的最优方案选择 ,即从若干决策方案中选择一个最优方案。决策问题中所含的方案数量在决策树中是隐性的。利用传统的决策树方法 ,无法确定一个次优方案 ,对所有方案进行排序也很难。从这个意义上来说 ,决策树法不如策略距阵法有效。该文结合一个实例 ,讨论了利用决策树对贝叶斯决策问题的决策方案进行选择和排序的方法。
Up to now, the decision tree method is used to find the optimal solution of a Bayesian decision problem, that is, to select an optimal one from several decision alternatives.The number of the alternatives of the decision tree is implicit.It is difficult to use the decision tree method to find the sub-optimal solution,not to mention the alternative ranking.In this sense, the decision tree method is less effective than the strategic matrix method. This paper discusses how to use the decision tree method to select and rank the alternatives. A practical case study is given to illustrate the applicability.
出处
《南京理工大学学报》
EI
CAS
CSCD
北大核心
2001年第5期450-452,共3页
Journal of Nanjing University of Science and Technology
基金
国家自然科学基金资助项目 (79870 0 30 )
关键词
决策分析
决策树
决策表
贝叶斯决策
排序
最优方案选择
医疗诊断
decision analysis, decision tree,decision table,Bayesian decision
alternative selection and ranking
