基于知识的决策评价函数  

A Knowledge Based Evaluating Function for Decision Making

作　　者：周青[1] 林剑文 黄志坚 

机构地区：[1]中山大学软件研究所,广东广州510275 [2]广州市秉政电子政务研究所,广东广州510033 [3]深圳市华成峰实业有限公司信息中心,广东深圳518057

出　　处：《管理科学与研究（中英文版）》2013年第2期16-22,共7页Management Science and Research

摘　　要：本文提出了一个决策评价函数，这个评价函数是通过对决策的目标、已经掌握的知识和证据进行分析而得到的。过去的经验和实验的结果作为样本在评价函数中有很好的使用；合格的样本可以从所有的样本中清楚地区分出来，这也是这个评价函数一个重要的好处。所有的有关概念都给出了清楚的定义，评价函数的构造有详细的描述。这个评价函数可以通过逻辑程序设计在计算机上实现并且可以在许多决策工具中使用，如机器学习，遗传算法、神经网络等。A decision evaluating function, proposed in this paper, is obtained from the analysis on the objectives, knowledge and evidence in our hand. Samples mainly of previous experimental results and experience have been fully utilized, of which the qualified can be distinguished; as the most noticeable merit of evaluating function. In this paper, the structure of evaluating function along with its relevant terms has been identified in details. The realization of evaluating function is through logical program set in computer, meanwhile, its applications are great, such as, machine learning, GA, neural network and so on.

关 键 词：决策评价函数 决策 基于知识 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]