arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 0912.4660cs.ITmath.IT
收藏

从指数族中寻找信息发散的最大化子

Finding the Maximizers of the Information Divergence from an Exponential Family

Johannes Rauh

摘要:

本文研究了指数族$E$的信息发散最大化问题。证明了极大化者$P$到$E$的$rI$ -投影是$P$和概率测度$P_-$的凸组合,该概率测度具有不相交的支持度和相同的充分统计量值$A$ .这个观察结果可以用来把原来的问题最大化$D(\cdot||E)$在所有的概率测量的集合到最大化的函数$\Dbar$在一个凸子集的$\ker A$。这两个问题的全局最大化者是相互对应的。此外,找到$\Dbar$的所有局部最大化器产生$D(\cdot||E)$的所有局部最大化器。本文还提出了两个算法来寻找$\Dbar$的最大化,并将其应用到两个例子,其中$D(\cdot||E)$的最大化是未知的。

英文摘要:

This paper investigates maximizers of the information divergence from an exponential family $E$. It is shown that the $rI$-projection of a maximizer $P$ to $E$ is a convex combination of $P$ and a probability measure $P_-$ with disjoint support and the same value of the sufficient statistics $A$. This observation can be used to transform the original problem of maximizing $D(\cdot||E)$ over the set of all probability measures into the maximization of a function $\Dbar$ over a convex subset of $\ker A$. The global maximizers of both problems correspond to each other. Furthermore, finding all local maximizers of $\Dbar$ yields all local maximizers of $D(\cdot||E)$. This paper also proposes two algorithms to find the maximizers of $\Dbar$ and applies them to two examples, where the maximizers of $D(\cdot||E)$ were not known before.

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。

↑