Optimal Distributed Goodness-of-Fit Testing for Discrete Distributions Under Privacy and Communication Constraints in the Large Sample Regime
In distributed goodness-of-fit testing for discrete distributions with large local sample sizes, leveraging statistical equivalence to Gaussian models reveals that the minimax rates for bandwidth and differential privacy constraints mirror those of the Gaussian case, highlighting the impact of data distribution on communication complexity.
基於覆蓋碼的近似最優量化器在獨立性分佈式檢測中的應用研究
本文提出了一種基於覆蓋碼的近似最優量化器，用於解決分佈式環境下針對獨立性的假設檢測問題，並通過分析錯誤概率的上下界，證明了其在短碼長和大碼長情況下的有效性。
분산 독립성 검정을 위한 준 최적 양자화기로서의 커버링 코드 (Covering Codes as Near-Optimal Quantizers for Distributed Testing Against Independence)
이진 선형 코드, 특히 최적의 커버링 반지름을 갖는 코드가 분산 가설 검정에서 독립성 검정을 위한 준 최적 양자화기로서 효과적으로 사용될 수 있음을 보여줍니다.
分散独立性検定のための準最適量子化器としての被覆符号：短い符号長と長い符号長における誤り確率の最適化
本稿では、二元線形符号を用いた分散型仮説検定における最適な量子化手法を提案し、短い符号長と長い符号長それぞれの場合において、誤り確率を最小化する手法を議論しています。
Near-Optimal Quantizers for Distributed Hypothesis Testing Against Independence Using Covering Codes
This research paper explores the use of binary linear codes, specifically those with optimal covering radius, as near-optimal quantizers in distributed hypothesis testing against independence for binary sources.
© 2024 by Linnk AI