摘要
对３种抽样方式讨论了余重对数二元回归概率模型的极大似然估计和极大惩罚似然估计。并对数据及模型作了进一步的推广。
篣nder three sampling schemes，the maximum and the penalized maximum likelihood esti-mation are disussed for the binary compiementary log-log regression probabilistic models．And wemake some extension to data and the models and made
关键词
似然估计
惩罚似然估计
余重对数模型
：iikelihood estimation，penalized likelihood estimation，binary complementary log-logmodels，generalized linear models.
