联邦离线强化学习数据集
用于联邦离线强化学习的多站点脓毒症数据集，支持个性化动态治疗策略优化。
基本信息
资源简介
联邦离线强化学习数据集
【一句话概览】
本数据集是 figshare 公开的结构化表格数据集，收录 4 个文件、合计 21.12 MB，主题涵盖 Dynamic treatment regimes、Electronic health records、Multi-source learning，可用于统计建模、特征工程与队列分析，CC BY 4.0 许可开放获取。
【数据集背景】
本数据集源站标题为《Federated Offline Reinforcement Learning》。配套论文《Federated Offline Reinforcement Learning》发表于 Journal of the American Statistical Association（2024），DOI：10.1080/01621459.2024.2310287。项目资助方：Tianxi Cai acknowledges the support of NIH (R01LM013614, R01HL089778). Junwei Lu acknowledges the support of NIH (R35CA220523, R01ES32418, U01CA209414). Zhaoran Wang acknowledges the support of NSF (Awards 2048075, 2008827, 2015568, 1934931), Simons Institute (Theory of Reinforcement Learning), Amazon, J.P. Morgan, and Two Sigma.。源站首次发布：2024-02-20。
【数据内容】
源站原始描述：Evidence-based or data-driven dynamic treatment regimes are essential for personalized medicine, which can benefit from offline reinforcement learning (RL). Although massive healthcare data are available across medical institutions, they are prohibited from sharing due to privacy constraints. Besides, heterogeneity exists in different sites. As a result, federated offline RL algorithms are necessary and promising to deal with the problems. In this article, we propose a multi-site Markov decision…
文件 README.md 为文本内容，开头为：# Federated Dynamic Treatment Regime (FDTR) 1. Supplementary This document contains the supplementary material to the paper ``Federated Offline Reinforcem…
【数据结构与技术规格】
文件数 4 个，合计 21.12 MB。README.md：text/plain，1.81 KB。Supplementary.pdf：application/pdf，304.05 KB。数据集 DOI：10.6084/m9.figshare.25134317.v2。
【主题与分类】
源站学科分类：Pharmacology、Biotechnology、Biological Sciences not elsewhere classified、Information Systems not elsewhere classified、Mathematical Sciences not elsewhere classified、Science Policy。主题标签：Dynamic treatment regimes、Electronic health records、Multi-source learning。
【适用方向】
结构化表格数据可直接用于统计建模、特征工程与队列分析。
【获取与许可】
源站页面：https://tandf.figshare.com/articles/dataset/Federated_Offline_Reinforcement_Learning/25134317
使用许可：CC BY 4.0（https://creativecommons.org/licenses/by/4.0/）。
规范引用：Federated Offline Reinforcement Learning. Taylor & Francis. Zhou, Doudou; Zhang, Yufeng; Sonabend-W, Aaron; Wang, Zhaoran; Lu, Junwei; Cai, Tianxi
【使用注意】
· 源站最后修改于 2024-04-01，引用时请注意版本时效。
下载信息
注册下载
需要注册 Kaggle 账号并登录后下载，适合需要跟踪下载记录和使用 API 的用户。
暂未开放公开下载
无需注册即可直接获取公开样本或文档，适合快速预览和评估数据集质量。
免登录有偿下载
公开数据集受托下载与技术交付服务。
提供高速下载与技术交付服务（收技术服务费，非数据销售）
暂未开放当前数据集主要来源为 Kaggle 公开托管，完整影像包建议通过原始链接或 Kaggle API 获取。
使用方式
数据集获取（Figshare）
- 源站标题：Federated Offline Reinforcement Learning
- DOI：10.6084/m9.figshare.25134317.v2
- 发布日期：2024-02-20
- 文件数：4 个
- 文件总体积：21.12 MB
- 源站页面：https://tandf.figshare.com/articles/dataset/Federated_Offline_Reinforcement_Learning/25134317
命令行下载
curl -L -o "Code.zip" "https://ndownloader.figshare.com/files/44375786"
curl -L -o "README.md" "https://ndownloader.figshare.com/files/44612656"
curl -L -o "Supplementary.pdf" "https://ndownloader.figshare.com/files/44612662"
curl -L -o "uasa_a_2310287_sm4432.docx" "https://ndownloader.figshare.com/files/44375795"
文件清单
Code.zip（20.79 MB）README.md（1.81 KB）Supplementary.pdf（304.05 KB）uasa_a_2310287_sm4432.docx（28.75 KB）
文件总体积 21.12 MB（源站 API 实测）。
数据说明
Evidence-based or data-driven dynamic treatment regimes are essential for personalized medicine, which can benefit from offline reinforcement learning (RL). Although massive healthcare data are available across medical institutions, they are prohibited from sharing due to privacy constraints. Besides, heterogeneity exists in different sites. As a result, federated offline RL algorithms are necessary and promising to deal with the problems. In this article, we propose a multi-site Markov decision process model that allows for both homogeneous and heterogeneous effects across sites. The proposed model makes the analysis of the site-level features possible. We design the first federated policy optimization algorithm for offline RL with sample complexity. The proposed algorithm is communication-efficient, which requires only a single round of communication interaction by exchanging summary statistics. We give a theoretical guarantee for the proposed algorithm, where the suboptimality for the learned policies is comparable to the rate as if data is not distributed. Extensive simulations demonstrate the effectiveness of the proposed algorithm. The method is applied to a sepsis dataset in multiple sites to illustrate its use in clinical settings. Supplementary materials for this article are available online including a standardized description of the materials available for reproducing the work.
许可
CC BY 4.0
引用
学术使用请引用 DOI 10.6084/m9.figshare.25134317.v2（source: https://tandf.figshare.com/articles/dataset/Federated_Offline_Reinforcement_Learning/25134317）。
数据缺失？
依托陕西、四川两大基地，我们与超过十家三甲医院建立直接合作关系，覆盖合规授权、采集治理、专业标注、数据交付的全流程，为AI医疗团队提供即拿即用的高质量临床数据。
⚡️ 需要数据支持或标注服务？立即联系我们获取专业方案。
