用神经网络法鉴别十二碳共轭双烯-1-醇、醛和乙酸酯双键位置异构体  

The Identification of Conjugated Bond Positional Isomers of Dodecadienic Compounds by Neural Networks

作　　者：蔡煜东[1] 甘骏人[1] 姚林声 

机构地区：[1]中国科学院上海冶金所,上海200050

出　　处：《有机化学》1992年第5期518-522,共5页Chinese Journal of Organic Chemistry

摘　　要：引言随着国际上神经网络研究的兴起,该法也逐步应用于化学模式识别领域。国内外很早就开展了化学模式识别的理论和应用研究,取得了很好的成果。最近,我们首次用神经网络进行了高温超导体制备条件的识别,成功率达100％。而后,又开展了生物制品(冠心病患者与健康人的血液样品)的多变量分析,利用神经网络探索微量金属元素与疾病(冠心病)The software to identify dodecadienic alcohols, aldehydes, acetetes and their doublebond positions by neural networks was developed, with which a group of samples were stu-died. The results indicate the reliability of neural networks. As compared with the fuzzymethod, the neural networks method is better.

关 键 词：神经网络 醇 醛 乙酸酯 质谱 

分 类 号：O623[理学—有机化学]