arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1703.09403math.STmath.PRstat.TH
收藏

奇异统计模型上的Cramér-Rao不等式I

The Cramér-Rao inequality on singular statistical models I

Jürgen Jost, Hông Vân Lê, Lorenz Schwachhöfer

摘要:

引入了参数化测度模型的本质切丛的概念和(可能奇异的)2-可积测度模型的约化Fisher度量的概念.使用这些概念和$k$ -可积参数化测度模型的新特征,我们将Cramér-Rao不等式扩展到$2$ -一般$φ$ -估计的可积(可能是奇异)统计模型,其中$φ$是$V$值特征函数,而$V$是拓扑向量空间。由此,我们在参数统计的最一般的通项条件下得到了一个内禀Cramér-Rao不等式.

英文摘要:

We introduce the notion of the essential tangent bundle of a parametrized measure model and the notion of reduced Fisher metric on a (possibly singular) 2-integrable measure model. Using these notions and a new characterization of $k$-integrable parametrized measure models, we extend the Cramér-Rao inequality to $2$-integrable (possibly singular) statistical models for general $φ$-estimations, where $φ$ is a $V$-valued feature function and $V$ is a topological vector space. Thus we derive an intrinsic Cramér-Rao inequality in the most general terms of parametric statistics.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑