arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1503.03613stat.MLcs.ITcs.LGmath.ITmath.PRmath.STstat.TH
收藏

论学习缺失的弥撒的不可能性

On the Impossibility of Learning the Missing Mass

University of Pennsylvania(宾夕法尼亚大学);University of California, Berkeley(加州大学伯克利分校);University of California, San Diego(加州大学圣地亚哥分校)

Elchanan Mossel, Mesrob I. Ohannessian

摘要:

本文表明,不施加进一步的结构性假设,就不能学习罕见事件的概率。关注事件是指获得i.i.d.覆盖范围之外的结果。从离散分布采样。该事件的概率被称为“缺失质量”。不可能的结果可以表述为:丢失的质量在相对误差中不是分布自由的PAC学习。证明是半建设性的,并依赖于一个耦合参数使用抖动几何分布。这一结果正式证实了民间传说,即为了预测罕见事件,人们必然需要具有“重尾”的分布。

英文摘要:

This paper shows that one cannot learn the probability of rare events without imposing further structural assumptions. The event of interest is that of obtaining an outcome outside the coverage of an i.i.d. sample from a discrete distribution. The probability of this event is referred to as the "missing mass". The impossibility result can then be stated as: the missing mass is not distribution-free PAC-learnable in relative error. The proof is semi-constructive and relies on a coupling argument using a dithered geometric distribution. This result formalizes the folklore that in order to predict rare events, one necessarily needs distributions with "heavy tails".

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。

↑