arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1504.06029cs.ITmath.IT
收藏

关于非渐近状态下量化观测值的MMSE估计

On MMSE estimation from quantized observations in the nonasymptotic regime

Jaeho Lee, Maxim Raginsky, Pierre Moulin

摘要:

本文研究了基于量化噪声观测的MMSE估计。它提出了非渐近界的MMSE遗憾,由于量化的两个设置:(1)估计的标量随机变量给定的量化矢量的$n$条件独立的意见,和(2)估计的$p$维随机矢量给定的量化矢量的$n$意见(不一定是独立的)时,完整的MMSE估计具有亚高斯浓度属性。

英文摘要:

This paper studies MMSE estimation on the basis of quantized noisy observations. It presents nonasymptotic bounds on MMSE regret due to quantization for two settings: (1) estimation of a scalar random variable given a quantized vector of $n$ conditionally independent observations, and (2) estimation of a $p$-dimensional random vector given a quantized vector of $n$ observations (not necessarily independent) when the full MMSE estimator has a subgaussian concentration property.

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。

↑