| 杨惠,陈利平,谢传欣,石宁,陈网桦.烃类沸点的定量构效关系研究[J].火灾科学,2011,20(1):62-67. |
| 烃类沸点的定量构效关系研究 |
| Quantitative structure-property relationships for boiling points of hydrocarbon compounds based on SVM |
| 投稿时间：2010-11-12 修订日期：2010-12-16 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI：10.3969/j.issn.1004-5309.年.期.顺序 |
| 基金项目:化学品安全控制国家重点实验室开放研究基金 |
|
| 中文关键词: 烃类物质 沸点 支持向量机(SVM) 定量构效关系(QSPR) |
| 英文关键词:Hydrocarbon compounds Boiling point Support vector machine(SVM) Quantitative structure-property relationship(QSPR) |
| 摘要点击次数: 749 |
| 全文下载次数: 799 |
| 中文摘要: |
| 应用CODESSA软件计算296种烃类物质的分子结构描述符,分别用启发式回归(HM)和最佳多元线性回归(B-MLR)筛选计算出的所有分子描述符,并建立沸点的线性回归模型。用B-MLR方法筛选出的4个描述符作为支持向量机(SVM)的输入建立了非线性模型。预测结果表明:所建立的模型稳健,泛化能力强,预测误差小。非线性模型(R2=0.9905,RMSE=10.2295)的性能优于线性回归模型(HM:R2=0.9819,RMSE=14.0606;B-MLR:R2=0.9842,RMSE=13.1058),预测效果令人满意。 |
| 英文摘要: |
| 296 molecular descriptors of hydrocarbon compounds were calculated by the CODESSA program,and these descriptors were pre-selected by heuristic method(HM) and best multi-linear regression method(B-MLR).Four-descriptor linear models were developed by the two methods to describe the relationship between the molecular structures and the boiling points.Using the four descriptors which were selected by B-MLR,the non-linear regression model was established based on the support vector machine(SVM).The predicted results indicated that the models had robustness,strong generative ability and small prediction error.The performance of the non-linear model(R2=0.9905,RMSE=10.2295) was better than that of the linear model(HM:R2=0.9819,RMSE=14.0606;B-MLR: R2=0.9842,RMSE=13.1058). |
| 关闭 |
