arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 1902.02890cs.ITcs.LGmath.ITmath.STstat.TH
收藏

基于Fisher信息的通信约束下学习分布的下界

Lower Bounds for Learning Distributions under Communication Constraints via Fisher Information

Stanford University(斯坦福大学)

Leighton Pate Barnes, Yanjun Han, Ayfer Ozgur

摘要:

我们考虑在分布式网络中学习高维、非参数和结构化(例如高斯)分布的问题,其中网络中的每个节点观察来自底层分布的独立样本,并且可以使用$k$位将其样本传送到中央处理器。我们考虑三种不同的通信模式。在独立模型下,每个节点通过独立地将其样本编码成$k$比特来将其样本传送到中央处理器。在更一般的顺序或黑板通信模型下,节点可以交互式地共享信息,但每个节点被限制为在最终转录本上最多写入$k$位。我们的特点的影响,通信约束$k$的极小极大风险估计的基础分布下$\ell^2$损失。我们开发的极大极小下界,适用于在一个统一的方式,许多常见的统计模型,并揭示了通信约束的影响可以是定性不同的,这取决于与每个模型相关联的得分函数的尾部行为。在我们的证明中的一个关键因素是量化样本的Fisher信息的几何特征。

英文摘要:

We consider the problem of learning high-dimensional, nonparametric and structured (e.g. Gaussian) distributions in distributed networks, where each node in the network observes an independent sample from the underlying distribution and can use $k$ bits to communicate its sample to a central processor. We consider three different models for communication. Under the independent model, each node communicates its sample to a central processor by independently encoding it into $k$ bits. Under the more general sequential or blackboard communication models, nodes can share information interactively but each node is restricted to write at most $k$ bits on the final transcript. We characterize the impact of the communication constraint $k$ on the minimax risk of estimating the underlying distribution under $\ell^2$ loss. We develop minimax lower bounds that apply in a unified way to many common statistical models and reveal that the impact of the communication constraint can be qualitatively different depending on the tail behavior of the score function associated with each model. A key ingredient in our proofs is a geometric characterization of Fisher information from quantized samples.

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。

↑