摘要
本文以多发性骨髓瘤的基因表达数据为例 ,利用数据挖掘技术 ,提出了一种针对基因表达数据进行知识发现的方法 .该方法通过计算基因的信息增益 ,结合神经网络 ,找出了特征基因集合 ,最后利用决策树进行特征规则的提取 ,给出了基于多发性骨髓瘤数据样本的产生式规则 ,为生物医学研究提供了一种分析和研究基因表达数据的参考方法 .实验结果表明了该方法的有效性 .
Based on the gene expression profiles of multiple myeloma,a method was proposed for knowledge discovery using data mining and machine learning methods.We used the information gain as the criterion of each gene for classification,and got a set of informative genes using artificial neural networks.Identified by decision tree algorithm,production rules were discovered as references for biomedical researchers.The effectiveness of the method we proposed is proved by experimental results.The method can also be used as a tool for gene expression analysis in the research of biomedicine and biotechnology.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2004年第9期1479-1482,共4页
Acta Electronica Sinica
基金
国家自然科学基金重点资助项目 (No .60 2 340 2 0 )
