Metraipolis › archive › dc_4d1ea650187b

Bandits with Probing: Optimal Regret and the Limits of Winner Feedback — arXiv:2609.15248

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.15248
License (as archived)
CC0-1.0
Archived
2026-10-16T00:11:29.077Z
SHA-256 of the archived text
a6edf8a6b1a5ca9bdd65f1704f7d55de4df9a8ee64a9d06ab18ba464c5a854a4
Stable URL of this page
https://metraipolis.com/documento/dc_4d1ea650187b

How to cite

Bandits with Probing: Optimal Regret and the Limits of Winner Feedback — arXiv:2609.15248 — arXiv (en), https://arxiv.org/abs/2609.15248 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_4d1ea650187b (Metraipolis archive dc_4d1ea650187b, sha256 a6edf8a6b1a5ca9bdd65f1704f7d55de4df9a8ee64a9d06ab18ba464c5a854a4)
BibTeX
@misc{dc4d1ea650187b,
  title = {Bandits with Probing: Optimal Regret and the Limits of Winner Feedback — arXiv:2609.15248},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.15248},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_4d1ea650187b; sha256 a6edf8a6b1a5ca9bdd65f1704f7d55de4df9a8ee64a9d06ab18ba464c5a854a4},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.15248. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A learner probes at most $k$ of $n$ arms each round, receives the maximum of their rewards in $[0,1]$, and competes with the best fixed arm. When does the probing advantage pay for learning? We determine two minimax laws. Under independent stochastic rewards with winner feedback (the maximum and a winning label), or on arbitrary fixed sequences given a single signed contrast between block maxima, the minimax regret has order $\Phi_{n,k}(T)=\min\{\frac{n-k}{n}T,\frac{n-k}{k}\}$, $2\le k<n$. Under winner feedback, both arbitrary joint i.i.d. rewards and fixed sequences have minimax regret of order $R_{n,k}(T)=\frac{n-k}{n}\min\{T,\frac{n+T}{k},\sqrt{\frac{nT}{k}}\}$. Both laws have universal constants and anytime upper bounds. The first reduces regret to a pure coverage cost: same-round contrasts absorb the stability cost, and independence permits exact resampling whose gains fund sample advancement. The second adds a learning cost that becomes comparable to coverage at horizon $n$; beyond $nk$, numerical maxima improve over labels alone. The lower bound allows every adaptive action size.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Data Structures and Algorithms (cs.DS); Machine Learning (stat.ML).
Pagina del lavoro: https://arxiv.org/abs/2609.15248 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-15T23:54:14.747Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).