Metraipolis › archive › dc_88cb8a525403

Odds-Ratio Thompson Sampling: A Specification and Design Guide for Contrast-Based Multi-Armed Bandits — arXiv:2609.19709

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.19709
License (as archived)
CC0-1.0
Archived
2026-10-10T04:08:42.950Z
SHA-256 of the archived text
af01f94ce27bbc574479d5111921db24b0a098aa893848ebe518cf9f60d81437
Stable URL of this page
https://metraipolis.com/documento/dc_88cb8a525403

How to cite

Odds-Ratio Thompson Sampling: A Specification and Design Guide for Contrast-Based Multi-Armed Bandits — arXiv:2609.19709 — arXiv (en), https://arxiv.org/abs/2609.19709 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_88cb8a525403 (Metraipolis archive dc_88cb8a525403, sha256 af01f94ce27bbc574479d5111921db24b0a098aa893848ebe518cf9f60d81437)
BibTeX
@misc{dc88cb8a525403,
  title = {Odds-Ratio Thompson Sampling: A Specification and Design Guide for Contrast-Based Multi-Armed Bandits — arXiv:2609.19709},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.19709},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_88cb8a525403; sha256 af01f94ce27bbc574479d5111921db24b0a098aa893848ebe518cf9f60d81437},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.19709. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Batched multi-armed bandits update on a service's own schedule, and the usual implementation carries each arm's absolute reward rate from one update to the next. When the shared level moves between batches, that memory goes stale even though the comparisons between arms may not have. Odds-Ratio Thompson Sampling (OR-TS) instead carries the joint posterior over log-odds contrasts and fits the common level afresh in every batch, marginalizing it out. This paper specifies that update, places it inside a Bayesian bandit agent with two controls, decay for how much past evidence survives an update and aggressiveness for how sharply belief becomes allocation, and evaluates it against absolute-rate memory. Across 86 public A/B series the level varies about twenty-five times more than the contrast. In prespecified synthetic environments a moving level costs absolute-rate memory five times the regret and leaves the best arm below a majority of traffic in 7 of 20 runs, against none for OR-TS. In a policy simulation built from 71 real experiments, where the contrasts are too small to resolve, expected-click differences stay within 0.1% for 58 of them, yet contrast memory still ends on the better arm more than twice as often. Where the contrasts themselves move, the bet fails, and that case is reported too.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.19709 — 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 2026-10-10T03:48:18.281Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).