节点文献
两类子样的估计
ESTIMATION FOR TWO TYPES OF SAMPLES
【摘要】 随机误差Δ是连续随机量 ,不可能严密地区分出正常误差和粗差 ,用抗差估计对粗差进行处理是不严密的 ,且对估计结果带来人为因素的影响。依随机误差所具有的连续性质进行估计的方法 ,克服了抗差估计方法的缺点 ,从而提高了估计质量 ,使估计结果既不纳伪、也不弃真。
【Abstract】 Random error Δ is a continuous variable,in which it is impossible to differentiate natural errors from gross ones rigorously. It is imprecise method to deal with gross error by robust estimation, and the estimation result will also be suffered factitiously. According to the continuous character of the random error,this paper propose an estimation method which overcomes the shortcoming of robust estimation and improve the quality of estimation. Thus the result of estimation does not include the false, nor does it forsake the true.
【关键词】 估计域;
理论平均值;
中位值;
极大似然估计;
概率平均值;
【Key words】 estimation field; theoretical average; median; maximum likelihood estimation; probability average;
【Key words】 estimation field; theoretical average; median; maximum likelihood estimation; probability average;
【基金】 广西自然科学基金资助项目 (批号 964 2 0 10 )
- 【文献出处】 大地测量与地球动力学 ,Crustal Deformation and Earthquake , 编辑部邮箱 ,2004年01期
- 【分类号】P315.75
- 【被引频次】3
- 【下载频次】62