摘要
论文对支持向量机(SVM)和正交设计方法进行了比较。支持向量机在分类和回归方面广为应用,而正交设计方法在实验设计方面是非常有效的,并且在化学工业中应用广泛。本文使用了两因素、七维正交实验(干燥实验)作为例子来把支持向量机方法应用到实验设计中去。正交表是用来研究实验的最优化条件和显著因素,本文给出了支持向量机和正交实验设计的计算结果。通过两者的比较,可以看到支持向量在实验预测方面比正交设计方法效果更好。从而可知支持向量机在实验设计方面的前景广阔。
The purpose of this paper is to provide a discussion and comparison of Support Vector Machine(SVM)and the orthogonaldesign method.SVM is a popular technique for classification and regression.Orthogonal test design is a scientific and effective methodin planning experiment,and it is extensively applied in chemistry industry.This paper,using 2 factors and 7 dimensions orthogonaltest(drying experiment)as an example,shows how to use the support vector machine to orthogonal test table and support vector ma-chine's applications on test design. The orthogonal test table was used to studying the optimal conditions of experiment,and the signifi-cant factors.The paper also gives the result of applying support vector machine to the testing result of orthogonal test design.By thecomparison,it has been proved that the prediction ability of support vector machine is more powerful than orthogonal design method inthe drying experiment.Support vector machine has a promising application in experimental design.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2004年第6期800-802,共3页
Computers and Applied Chemistry
基金
Supported by the LiuHui Center for Applied Mathematics
Nallkai University and Tianjin University
关键词
支持向量机(SVM)
正交设计
核函数
support vector machine(SVM)
orthogonal design
kernel function
