论文部分内容阅读
符号推理系统,或者说基于规则的专家系统,已成功地应用于从过程控制、信用评价到保险证券包销和医疗诊断等众多领域。但是,业已证明仍有一些问题仅仅使用纯符号技术是难以解决的。在许多这类领域中,基于人工神经网络,即连接机制的系统大有希望,从而产生了一个新的研究领域——混合型系统。在 Elaine Rich 领导下,MCC 公司的 AI实验室正着手进行一个混合型系统的开发项目,可望在91年内研制出一个用于过程控制的实用模型系统。Rich 曾在布朗大学获语言
Symbolic reasoning systems, or rule-based expert systems, have been successfully used in many areas ranging from process control, credit rating to underwriting of insurance securities and medical diagnostics. However, it has been proven that there are still some problems that are difficult to solve using purely symbolic techniques. In many of these areas, promising systems based on artificial neural networks, known as connection mechanisms, have led to a new area of research - hybrid systems. Under Elaine Rich, MCC’s AI lab is embarking on a hybrid system development project that is expected to develop a practical model system for process control within 91 years. Rich has a language at Brown University