Logistic模型的统计诊断  

Statistical Diagnosis of Logistic Model

作　　者：光琳[1] 宗序平[2] 

机构地区：[1]江苏联合职业技术学院连云港财经分院,江苏连云港222003 [2]扬州大学数学科学学院,江苏扬州225002

出　　处：《江南大学学报（自然科学版）》2012年第1期113-117,共5页Joural of Jiangnan University (Natural Science Edition)　

摘　　要：系统地分析了Logistic模型的统计诊断,将线性模型的诊断方法运用到Logistic模型。推导出数据删除模型一阶近似的参数估计公式,并给出数据删除模型与均值漂移模型的等价性;推导了判别强影响点或异常点的诊断统计量如Score统计量、广义Cook距离、似然距离等。通过实例分析证实了该诊断方法的有效性。This paper systematically discusses the statistical diagnostics of Logistic statistical model, which are extended from the diagnostic methods for linear model. The first-order approximate formula of parameter estimate of CDM are derived,and the equivalence of CDM and MSOM are proved. At the same time, we derivate the diagnostic statistics to distinguish strong influence points or the diagnosis of outliers, such as Score statistics,Generalized Cook distance, and Likelihood distance. Finally, a case study is given to illustrate our results.

关 键 词：LOGISTIC模型 统计诊断 诊断统计量 

分 类 号：O212.1[理学—概率论与数理统计]