基于支持向量机的miRNA预测方法研究  

Research of miRNA Prediction Based on SVM

作　　者：张祖勇[1] 孙龙清[1] 

机构地区：[1]中国农业大学,100083

出　　处：《中国科技信息》2008年第20期43-45,共3页China Science and Technology Information

摘　　要：本文从miRNA及其前体的生物学特征出发,在对支持向量机理论及其应用特点进行研究的基础上,构建了基于支持向量机的miRNA预测过程模型,在miRNA特征的向量表示、miRNA特征选择、预测模型核函数及参数选择方面进行了研究。以水稻、拟南芥、玉米的miRNA为实例,对基于支持向量机的miRNA预测方法的预测准确率进行了验证,实验结果表明该方法预测准确率达95%以上。According to the biological feature of miNRA and pre-miRNA, and based on the research of Support Vector Machine（SVM） theories aid application, a miRNA prediction process module is given in this thesis. Some attempt and research about vector denotionation of miRNA feature, miRNA features seiection, kereel function and parameters selection is done. An example with rice, arabidopsis thaliana. and corn is given; some research during miRNA prediction process is done. The efficiency of the miRNA predication module based on SVM is verified, and the accuracy race is more than 95%.

关 键 词：支持向量机 MIRNA 机器学习 

分 类 号：Q75[生物学—分子生物学]