摘要
本文提出了一种新的变量筛选法——正交递归选择法,该法可以得到预报能力较强的模型,即PRESS(预报残差平方和)值较低的模型.用该法处理构效关系问题,并与逐步回归正向选择法及PLS回归法进行了比较,得到满意的结果。
A new model selection method named orthogonalization recurrence selection (ORS) method is proposed. By use of this method the model with higher predictive ability can be ob-tainded or the lower PRESS statistic values of the model can be achieved. The comparison is made among forward selection method in the stepwise regression and PLS as well as ORS methods through the example of QSAR(quantitative structure activity relationship).
出处
《高等学校化学学报》
SCIE
EI
CAS
CSCD
北大核心
1993年第11期1518-1521,共4页
Chemical Journal of Chinese Universities
基金
国家人事部博士后基金
关键词
变量筛选
正交递归选择
逐步回归
Model selection, PRESS criterion, Stepwise regression, PLS regression, Gram-Schmidt ortllogonalization