| 刘遵雄,田珊珊.基于似然深度的指数分布的参数估计[J].井冈山大学自然版,2017,(5):20-23 |
| 基于似然深度的指数分布的参数估计 |
| ESTIMATOR BASED ON LIKELIHOOD DEPTH WITH APPLICATION TO THE PARAMETER OF EXPONENTIAL DISTRIBUTION |
| 投稿时间：2016-07-21 修订日期：2016-10-16 |
| DOI：10.3969/j.issn.1674-8085.2017.05.005 |
| 中文关键词: 有污染的截尾样本 指数分布 极大似然估计 似然深度估计 |
| 英文关键词: contaminated and truncated data exponential distribution maximum likelihood estimation likelihood depth estimation |
| 基金项目:国家自然科学基金项目（71361009） |
|
| 摘要点击次数: 1679 |
| 全文下载次数: 2655 |
| 中文摘要: |
| 针对传统估计方法如极大似然估计对于服从指数分布且有污染的截尾数据的参数估计并不是很理想的问题，提出使用一种新的估计方法对其进行参数估计，即似然深度估计，并通过两组实验进行对比，结果显示利用似然深度估计方法得到的参数偏差和均方差较小，表明似然深度估计是一种估计服从指数分布且有污染的截尾数据参数的有效方法。 |
| 英文摘要: |
| According to the traditional estimation methods such as maximum likelihood estimation (MLE) is not very ideal for estimating the parameter of the contaminated and truncated data following exponential distribution. A new estimation method is proposed to estimate the parameter of exponential distribution with the contaminated and truncated data, which is likelihood depth estimation (LDE). The result shows that the deviation and the mean square error (MSE) of LDE are the lower when compared to MLE through conducting two random simulation experiments, which indicates that LDE is an effective method for estimating the parameter of exponential distribution with the contaminated and truncated data. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
