摘要
针对传统专家系统获取知识困难自学习能力差和推理能力弱等缺点,设计并实现了基于BP神经网络的水牛疾病诊断专家系统。系统用专家以往诊断水牛疾病的病历来训练神经网络,并通过训练过的神经网络来实现疾病的诊断,然后针对诊断结果进行反向推理以确诊水牛疾病,结果表明,该系统很好地改进了传统专家系统存在的一些缺点。另外,该水牛疾病诊断系统的实现方法可推广应用到其它动物的疾病诊断系统中,为疾病诊断系统的开发提供了一条有效的新途径。
Aimed at the drawbacks of traditional expert system such as difficulty in obtaining knowledge,low ability of self-learning and inference,the carabao disease diagnosis system is designed and realized based on BP artificial neural network.The diagnosis system used the old cases to training the neural network,and the disease diagnosis is achieved through the trained neural network,then the back-ward inference is carried out so that the diagnose disease result is gotten ultimately.The results proved that the system improved some shortcomings of traditional expert systems.This disease diagnosis system could be popularized to the diagnosis system of other animals also.It provided a new effective method to the development of disease diagnosis system.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第6期1485-1488,共4页
Computer Engineering and Design
关键词
疾病诊断系统
反向传播神经网络
专家系统
训练样本
反向推理
disease diagnosis system
back propagation neural network
expert system
training specimen
backward inference
