刘杰杰,李霓*.基于估计方程的简单计数数据统计分析[J].海南师范大学学报自科版,2018,31(3):306-312 |
基于估计方程的简单计数数据统计分析 |
Statistical Analysis of Simple Count Data Based on Estimating Equations |
|
DOI:10.12051/j.issn.1674-4942.2018.03.010 |
中文关键词: 简单计数数据 估计方程 相合性 渐近正态性 |
英文关键词: simple count data estimating equations consistency asymptotic normality |
基金项目:海南省自然科学基金面上项目(117116,118MS042);国家自然科学基金项目(11861030,11471135) |
|
摘要点击次数: 1263 |
全文下载次数: 1138 |
中文摘要: |
当泊松分布和负二项分布无法满足简单计数数据分析要求时,建立一般的条件期望模
型并提出基于估计方程的参数估计方法,推导参数估计的渐近性质,包括估计的相合性和渐近正
态性。还通过数值模拟验证了这些估计在有限样本下的有效性。最后,将模型及估计方法应用于
一组临床试验记录的计数数据的分析中。结果表明,模型分析结果具有统计学意义且与试验情况
一致,具有医学应用价值。 |
英文摘要: |
When Poisson distribution and negative binomial distribution assumption cannot meet the requirement for sim⁃
ple count data analysis, a general conditional expectation model was built and a parameter estimation method based on the
estimating equation was proposed. The asymptotic properties of parameter estimation including consistency and asymptotic
normality was deduced. The effectiveness of these estimations with limited samples was verified by means of numerical sim⁃
ulations. At last, the model coupled with the estimation method was applied for the count data analysis of a set of clinical tri⁃
al record. The results indicated astatistically significance for the data analysis and were in accordance with the trials, which
indicated an application value in medicine. |
查看全文
查看/发表评论 下载PDF阅读器 |
关闭 |