摘要
建立了基于最大间隔分类的多环芳烃致癌性预测模型,实验表明该模型不仅具有较高的准确率,同时算法简单,泛化能力强,能够用于多环芳烃致癌性的分类和预测.
A predicting model of polycyclic aromatic hydrocarbons carcinogenic properties based on maximal margin classification is established, Experiments show that this model not only has a higher accuracy but also is simple in algorithm, generalizd in ability; and it can be used for polycyclic aromatic hydrocarbons carcinoge.nic classification and forecasting.
出处
《三峡大学学报（自然科学版）》
CAS
2008年第4期83-85,共3页
Journal of China Three Gorges University：Natural Sciences
关键词
多环芳烃
分类
间隔
二次规划
polycyclic aromatic hydrocarbons
classification
margin
quadratic programming
