论文部分内容阅读
本文提出一类计算机辅助诊断系统。它是基于相当数量和权威性的知认,通过特征抽取逐步限定的优化决策过程以代替或协助专家解决问题的计算机程序系统,此系统使用基于不确定信息逻辑,证据理论和FOX关于扩展概率框架的一种非单调逻辑确立了由商空间的优化缩影形成数量较少的判断规则(极小指征集族)的数学模型和作为似然推理模式的“相似性原则”。使用了具有较强的PROLOG自动推理能力的语言以表述知识,形成智能数据库并通过自动发现问题和与专家配合而动态地完善规则集得以实现学习功能。本系统在人工心脏起搏器的安植、更换中的成功应用,为推广、移植到其它领域展示了诱人的前景。
This paper presents a type of computer-aided diagnosis system. It is a computer program system based on a considerable amount and authoritative knowledge to extract the step-by-step optimization decision-making process by feature extraction in order to replace or assist the expert to solve the problem. The system uses uncertain information logic, evidence theory and FOX A non-monotonic logic establishes a mathematical model of a small number of judgment rules (minimal set of signatures) formed by the optimization of quotient space and the “principle of similarity” as a model of likelihood inference. A language with strong PROLOG automatic reasoning is used to present knowledge, form a smart database and fulfill learning by dynamically discovering problems and working with experts to dynamically refine rule sets. The system in the artificial heart pacemaker placement, replacement of the successful application of the promotion, transplantation to show other attractive prospects.