considered to be a founding father of Russian mathematics. Among his well-known students were the mathematicians Dmitry Grave, Aleksandr Korkin, Aleksandr Lyapunov, and Andrei Markov.
Chebyshev Inequality
例:指数分布参数的矩估计
设 为来自总体 的样本,求 的矩估计.
解:
因为只有一个待估参数,故只需令 .
对于指数分布,, 进而可得 或 .
故所求 的矩估计为 .
例:Gamma 分布参数的矩估计
设 为来自总体 的样本,求 的矩估计.
分析: 有两个待估参数,故令 , .
对于 ,, .
注: 因为 ,,可利用 替代 .
解: 由矩估计法,令
解之可得
,
使用矩估计法时,可以使用总体方差和修正的样本方差替换总体和样本的二阶矩.
例:均匀分布参数的矩估计
求均匀分布 中参数 的矩估计.
, .
令 , 解得 .
实验验证
以下为来自总体 的一组观测值
矩估计值 .
更“合理”的答案似乎应该是 , .
22.2 最大似然估计
Early users of maximum likelihood were C.F. Gauss, P.S. Laplace, etc.
However, its widespread use rose between 1912 and 1922 when Ronald Fisher recommended, widely popularized, and carefully analyzed maximum-likelihood estimation.
R.A. Fisher
Sir Ronald Aylmer Fisher (1890-1962), a English geneticist and statistician
"a genius who almost single-handedly created the foundations for modern statistical science"
"the single most important figure in 20th century statistics"