...
首页> 外文期刊>IEEE Transactions on Information Theory >Minimum KL-Divergence on Complements of $L_{1}$ Balls
【24h】

Minimum KL-Divergence on Complements of $L_{1}$ Balls

机译:$ L_ {1} $个球的补码的最小KL散度

获取原文
获取原文并翻译 | 示例
           

摘要

Pinsker's widely used inequality upper-bounds the total variation distance $leftVert P-QrightVert_{1}$ in terms of the Kullback–Leibler divergence $D(PVert Q)$ . Although, in general, a bound in the reverse direction is impossible, in many applications the quantity of interest is actually $D^{ast}({v},Q)$ —defined, for an arbitrary fixed $Q$ , as the infimum of $D(PVert Q)$ over all distributions $P$ that are at least ${v}$ -far away from $Q$ in total variation. We show that $D^{ast}({v},Q)leq C{v}^{2}+O({v}^{3})$ , where $C=C(Q)={1}/{2}$ for balanced distributions, thereby providing a kind of reverse Pinsker inequality. Some of the structural results obtained in the course of the proof may be of independent interest. An application to large deviations is given.
机译:Pinsker广泛使用的不等式将总变化距离$ leftVert P-QrightVert_ {1} $的上限限定为Kullback-Leibler散度$ D(PVert Q)$。尽管通常来说,反向限制是不可能的,但是在许多应用中,感兴趣的数量实际上是$ D ^ {ast}({v},Q)$ —对于任意固定的$ Q $,定义为至少$ {v} $的所有分配$ P $的$ D(PVert Q)$的最小值-远不及$ Q $的总变化量。我们显示$ D ^ {ast}({v},Q)leq C {v} ^ {2} + O({v} ^ {3})$,其中$ C = C(Q)= {1} / {2} $用于平衡分布,从而提供一种反向的Pinsker不等式。在证明过程中获得的一些结构性结果可能与个人利益无关。给出了大偏差的应用。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号