摘要
给出双向联想记忆 (BAM)神经网络的基本原理。在此基础上 ,根据血清中微量元素的含量 ,将双向联想记忆神经网络用于正常人与肺癌患者的分类判别。实验结果表明 :用独立预测样本作检验 ,在本工作所选定的条件下 ,可以达到 10 0 %的正确识别率。并讨论了双向联想记忆神经网络的影响因素。
The basic theory of bidirectional associative memory (BAM) was given and the BAM approach was applied to the classification of normal people and lung cancer patients based on the microelement contents in serum samples. This method was verified with independent prediction samples and ran discriminate all lung cancer patients at rate of 100 percent under certain conditions. The results obtained by BAM are slightly better than those obtained by back-propagation neural network (97%). Results showed that the BAM could be a promising aiding method for diagnosis of lung cancer. The effects of network parameters were investigated and related problem, were discussed.
出处
《光谱学与光谱分析》
SCIE
EI
CAS
CSCD
北大核心
2002年第2期341-344,共4页
Spectroscopy and Spectral Analysis
基金
国家教育部中青年骨干教师基金资助
关键词
双向联想记忆
人工神经网络
肺癌
微量元素
分类判别
血清
bidirectional associative memory
artificial neural network
classification
microelement
serum
lung cancer
