摘要
介绍了主成分的概念以及Logistic回归分析和支持向量机方法,建立了诊断脾虚证的Logistic回归模型和支持向量机模型.在这两个数学模型基础上设计了一个计算机辅助诊断脾虚证的原型系统.通过对测试病例所作的预测,显示在小样本情况下,支持向量机具有较好诊断效果.
Concepts of the principal components and approaches of the Logistic regression analysis and the support vector machines are introduced.Then two different mathematical models of spleen feebleness diagnosis are established.Based on these models,a prototype of computer-aided spleen feebleness diagnosis is designed.Through predictions made on clinical test samples,it is shown that the model of support vector machines behaves better in the cases of small amounts of samples.
出处
《暨南大学学报(自然科学与医学版)》
CAS
CSCD
北大核心
2006年第3期363-367,共5页
Journal of Jinan University(Natural Science & Medicine Edition)
基金
国家自然科学基金重点资助项目(90209004)
广东省自然科学基金资助项目(021149)
关键词
脾虚证诊断
LOGISTIC回归分析
支持向量机
数学模型
spleen feebleness diagnosis
Logistic regression analysis
support vector machines
mathematical models