PENALIZED MAXIMUM-LIKELIHOOD ESTIMATION IN LOGISTIC-REGRESSION AND DISCRIMINATION
PENALIZED MAXIMUM-LIKELIHOOD ESTIMATION IN LOGISTIC-REGRESSION AND DISCRIMINATION
复制标题
DOI：
10.1093/biomet/69.1.123
复制
发表时间：
1982-01-01
期刊：
影响因子：
2.7
通讯作者：
BLAIR, V
中科院分区：
文献类型：
--
作者：
ANDERSON, JA;BLAIR, V
二元logistic回归模型参数的极大似然估计|x）分别讨论了从（i）Hgivenx的条件分布，（ii）Handx的联合分布，和（iii）xgivenH的条件分布的抽样。与连续在后者的抽样方案的困难进行了讨论。为了避免这些，惩罚最大似然估计的介绍，这给估计的logistic参数和非参数样条估计的边缘分布的x。延伸到多项式逻辑回归概述。
Maximum likelihood estimation of the parameters of the binary logistic regression model for pr(H|x) is discussed with separate discussion of sampling from (i) the conditional distribution ofHgivenx, (ii) the joint distribution ofHandx, and (iii) the conditional distribution ofxgivenH. Difficulties associated with continuousxin the latter sampling scheme are discussed. To avoid these, penalized maximum likelihood estimates are introduced, which give estimates of the logistic parameters and a nonparametric spline estimate of the marginal distribution ofx. Extensions to multinomial logistic regression are outlined.