跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

让下一篇重要论文,不再错过。

覆盖计算机全学科,以 AI 为快捷入口。关注方向、作者和关键词,把发现变成持续的研究积累。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1072新增论文
+845其中 AI 相关
+94已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描:2026-11-09 19:37 · 全站最近入库:2026-11-09 13:22各信源进度 →

统计口径与数据新鲜度

统计 2026-11-08 20:28 — 2026-11-09 20:28(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-09 19:37最近发现新增版本:2026-11-09 13:22库内论文最新更新时间:2026-11-07 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

18,172 已收录论文15,506 AI 相关4,527 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 3 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-09 19:35(北京时间)
新论文 · v1arXiv:2611.09806

An Efficient Synchronization Scheme for Distributed Bandits

Raymond Zhang, Henrique Miyamoto, Richard Combes, Sheng Yang

原文摘要

We study the problem of synchronizing two players in a \emph{distributed multi-armed bandit} setting. The probability of \emph{synchronization error} is analyzed by recasting synchronization as a \emph{communication problem}. Using a well-designed random coding ensemble, we derive error exponents for the maximum-likelihood decoder. We then introduce a \emph{universal decoder} that achieves the same error exponents as the optimal decoder. Finally, applying these results to distributed multi-armed bandits, we show an upper bound on the regret of our synchronization scheme.

cs.IT

版本更新 · v2arXiv:2502.17175

Linear Bandits on Ellipsoids: Minimax Optimal Algorithms

Raymond Zhang, Hedi Hadiji, Richard Combes

原文摘要

We consider linear stochastic bandits where the set of actions is an ellipsoid. We provide the first known minimax optimal algorithm for this problem. We first derive a novel information-theoretic lower bound on the regret of any algorithm, which must be at least $Ω(\min(d σ\sqrt{T} + d \|θ\|_{A}, \|θ\|_{A} T))$ where $d$ is the dimension, $T$ the time horizon, $σ^2$ the noise variance, $A$ a matrix defining the set of actions and $θ$ the vector of unknown parameters. We then provide an algorithm whose regret matches this bound to a multiplicative universal constant. The algorithm is non-classical in the sense that it is not optimistic, and it is not a sampling algorithm. The main idea is to combine a novel sequential procedure to estimate $\|θ\|$, followed by an explore-and-commit strategy informed by this estimate. The algorithm is highly computationally efficient, and a run requires only time $O(dT + d^2 \log(T/d) + d^3)$ and memory $O(d^2)$, in contrast with known optimistic algorithms, which are not implementable in polynomial time. We go beyond minimax optimality and show that our algorithm is locally asymptotically minimax optimal, a much stronger notion of optimality. We further provide numerical experiments to illustrate our theoretical findings.

stat.ML · cs.LG

新论文 · v1arXiv:2611.09466

Symmetry Breaking over Multiway Channels

Raymond Zhang, Richard Combes, Sheng Yang

原文摘要

Identical users running the same algorithm must often break their symmetry and agree on distinct roles, as in leader election, naming, and resource allocation. We study anonymous symmetry breaking over a discrete memoryless multiway channel invariant under permutations of its $J$ users: each user observes only a private output and must choose a number in $\{1,\dots,J\}$, and the symmetry is broken if the $J$ numbers are distinct. Using a Rényi change of measure, we derive a finite-blocklength converse on the error probability of every symmetric causal protocol, including adaptive and randomized ones. For achievability, block coding with random names reduces the problem to all-source multicast, for which we derive a Gallager random-coding exponent for every fixed $J$, and a cut-set bound that limits the exponent of the probability that two names coincide or are decoded wrongly. Interactive role controllers instead let the users contend for identifiers directly: we give their exact error probability, an achievable exponent given by a spectral radius when some outputs certify a role, and a floor for finite-state controllers when every output has positive probability, and we introduce generalized musical chairs. On a noisy bandit channel motivated by multiplayer multi-armed bandits, with noiseless collisions and arms of a common mean below one, interactive schemes provably exceed the limit of name decoding, and numerical estimates suggest the same with noisy collisions. For two users, noiseless collisions, one informative arm and a silent arm, the optimal error probability is determined up to a factor of two.

cs.IT