English(EN) What a Data Release Must Preserve for Off-Policy Evaluation

新研究详细介绍了离轨评估的数据保留方法

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-14 04:00

arXiv上的一篇新论文介绍了一种通过表征数据发布中必须保留的区别来改进机器学习中离轨评估的方法。该研究详细介绍了如何通过在发布记录中保持特定动作的评估权重恒定并将历史记录按这些权重分组来确保评估的有效性。提出的修复算法旨在减小平均成对锐度宽度，并在24个KuaiRec任务的测试中显示出显著的减小。 AI

影响为提高机器学习评估的可靠性提供了理论框架，尤其是在数据有限或经过处理的情况下。

排序理由关于统计机器学习主题的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv stat.ML 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 3 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 关于统计机器学习主题的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv stat.ML TIER_1 English(EN) · Kyungbok Lee, Yating Zou, Michael R. Kosorok · 2026-10-14 04:00

数据发布在离线策略评估中必须保留什么

arXiv:2610.15283v1 Announce Type: new Abstract: In off-policy evaluation, merging decision-time histories that matter differently for a specified policy value or comparison can make that evaluation unidentified, even with observed actions and outcomes and known behavior and targe…

本文如何被排名

报道来源 [1]

数据发布在离线策略评估中必须保留什么

相关实体

相关话题