摘要
本文提出了一个规则可信度的学习算法,并设计了有关这个算法的适用于深度优先推理策略、宽度优先推理策略的仿真实验。实验结果表明这个学习算法可以有效地修正规则的可信度。目前,这个学习算法已用于一个工业专家系统。
This paper presents a learning algorithm for modification of rule confidences,and designs simulationexperiments for it.This learning algorithm as shown by simulation results can effectively moditfy the con-fidences of the rules in depth-first strategy and breadth-first strategy.The proposed learning algorithmhas been applied to an industrial expert system at present.
出处
《机器人》
EI
CSCD
北大核心
1989年第4期36-39,共4页
Robot
关键词
规则可信度
专家系统
优先推理
learning
rule confidence
depth-first
breadth-first
