基于支持向量机回归与K-最近邻法的组合预测用于除草剂QSAR建模

谭泗桥; 袁哲明; 柏连阳; 熊洁仪

基于支持向量机回归与K-最近邻法的组合预测用于除草剂QSAR建模

1.
湖南农业大学, 生物安全科学技术学院, 长沙 410128
2.
湖南农业大学, 信息科学技术学院, 长沙 410128
3.
湖南农业大学,生物安全科学技术学院, 长沙 410128

基金项目:

国家自然科学基金(30570351);教育部新世纪优秀人才支持计划项目.

计量
- 文章访问数: 1366
- HTML全文浏览量: 29
- PDF下载量: 398
- 被引次数: 0
出版历程
- 收稿日期: 2007-08-05

The QSAR Modeling of Herbicide Using a Combinatorial Model Based on Support Vector Regression and K-nearest Neighbor

1.
Bio-safety Science and Technology College,Hunan Agricultural University,Changsha 410128,China
2.
College of Information Science and Technology,Hunan Agricultural University,Changsha 410128,China

摘要

摘要: 为了提高定量构效关系(QSAR)研究的预测精度,发展了一种基于支持向量机回归(SVR)与K-最近邻法(KNN)的组合预测方法:以均方误差(MSE)最小为原则,基于SVR实施核函数寻优;基于MSE最小原则与最优核函数以SVR进行描述符筛选并得到保留描述符;通过"多轮末尾强制淘汰法"揭示各保留描述符对预测精度的影响程度;从保留描述符出发,以不同KNN预测值反映样本集异质性并构建子模型,然后基于SVR以留一法实施组合预测。运用该组合预测方法研究苯乙酰胺类除草剂QSAR建模,结果表明:基于SVR与KNN的组合预测方法在参比模型中预测精度最高,具结构风险最小、非线性、能有效克服过拟合、泛化推广能力优异等优点,在QSAR研究中具有广泛的应用前景。
- 支持向量机回归 /
- K-最近邻法 /
- 组合预测 /
- 定量构效关系
Abstract: To improve the precision of QSAR modeling,a novel nonlinear combinatorial forecast method based on support vector regression (SVR) and K-nearest neighbor (KNN) was proposed.Kernels and descriptors optimization based on SVR were evaluated by the rule of minimum MSE value,and "multi-round enforcement optimization" was taken to illuminate the effect of retained descriptors on forecasting precision.The heterogeneity of sample set was characterized by different KNN and multiple sub-models were assembled,then combinational forecast was carried out based on leave-one-out method.The new method had been employed to study for QSAR on a series of herbicidal materials,N-phenylacetamides,and has the highest prediction precision in all reference models.It has the advantages of structural risk minimization,non-linear characteristics,avoiding the over-fit and strong generalization ability,etc.The novel combination model,so,can be widely used in QSAR.
- support vector regression /
- K-nearest neighbor /
- combinatorial forecasting /
- QSAR

HTML全文

参考文献(18)

ZHONG Guo-hua(钟国华),HU Mei-ying(胡美英).QSAR及其在农药设计中的应用和进展[J].Chin J Pestic Sci(农药学学报),2001,3(2):1-11.

YIN Jia-jian(印家健),LI Meng-long(李梦龙),WEN Zhi-ning(文志宁).支持向量回归用于氨基酸描述符在肽QSAR建模中的性能评价[J].J Sichuan Univ (Nat Sci Ed)(四川大学学报,自然科学版),2006,43(2):396-402.

ZHOU Peng(周鹏),ZENG Hui(曾晖),LI Bo(李波).支持向量机分类和回归用于肽的QSAR研究[J].Chemistry(化学通报),2006,69(5):342-346.

DING Jun-jie(丁俊杰),DING Xiao-qin(丁晓琴),ZHAO Li-feng(赵立峰).新型三维氨基酸结构描述符的研究及其在多肽QSAR中的应用[J].Acta Pharmaceutica Sinica(药学学报),2005,40(4):340-346.

AI Na(艾娜),WU Zuo-wei(吴作伟),REN Jiang-hua(任江华).支持向量机与人工神经网络[J].J Shandong Univ Tech (Sci & Tech)(山东理工大学学报,自然科学版),2005,19(5):45-49.

VAPNIK V.The Nature of Statistical Learning Theory [M].New York: Springer Verlag Press,1995: 87-189.

CRISTIANINI N,SHAWE-TAYLOR J.An Introduction to Support Vector Machines and other Kernel-based Learning Methods(支持向量机导论)[M].LI Guo-zheng(李国正,等译).Beijing(北京):Publishing House of Electronic Industry(电子工业出版社),2004:82-139.

DENG Nai-yang(邓乃扬),TIAN Ying-jie(田英杰).The New Method of Data Mining-Support Vector Machine(数据挖掘中的新方法——支持向量机)[M].Beijing(北京):Science Press(科学出版社),2004.

BATES J M,GRANGER C W J.Combination of Forecasts[J].Operations Research Quarterly,1969,20(4):451-468.

LI Jian(李剑),CHEN De-zhao(陈德钊).优化的径向基-循环子空间网络为药物定量构效关系建模[J].Chin J Anal Chem(分析化学),2005,33(6):767-771.

TIAN Sheng-feng(田盛丰),HUANG Hou-kuan(黄厚宽).支持向量机回归的简化算法[J].Journal of Software(软件学报),2002,13(6):1169-1172.

WANG Xiao-ye(王晓晔),WANG Zheng-ou(王正欧).K-最近邻分类技术的改进算法[J].Journal of Electronics & Information Technology(电子与信息学报),2005,27(3):487-491.

ZHANG Guo-ping(张国平).B-G组合预测剖析[J].Forecasting(预测),1988,12(5):24-27.

CAI Gen-xiang(柴根象),WU Yue-qin(吴月琴).相依样本下污染线性模型的最近邻估计[J].Acta Mathematicae Applicatae Sinica(应用数学学报),2006,29(3):542-554.

LI Yuan-cheng(李元诚),LI Bo(李波),FANG Ting-jian(方廷健).基于小波支持向量机的非线性组合预测方法研究[J].Information and Control(信息与控制),2004,33(3):303-306.

TANG Qi-yi(唐启义),FENG Guang-ming(冯明光).DPS Data Processing System for Practical Statistics(实用统计分析及其DPS数据处理系统)[M].Beijing(北京):Science Press(科学出版社),2002:579-585.

CHEN Y Q,CHEN D Z,HE C Y.Quantitative Structure-activity Relationships Study of Herbicides Using Neural Networks and Different Statistical Methods[J].Chemometrics and Intelligent Laboratory Systems,1999,45:267-276.

KIRINO O,TAKAYAMA C,MINE A.Quantitative Structure- activity Relationships of Herbicidal N-(1-Methyl-1-phenylethyl) phenylacetamides[J].J Pestic Sci,1986,11:611-617. (Ed.JIN S H)

施引文献

资源附件(0)

访问统计