Federated Learning with Non-IID Data
- 类型:arxiv
- 标识:1806.00582
- 链接:https://arxiv.org/abs/1806.00582
- 主题:evaluation
- 主分类:engineering
- 形态:method
- 被引:3417
- 被引来源:Semantic Scholar
- S2被引:3417
- OpenAlex被引:1917
- 影响力被引:296
- TLDR:This work presents a strategy to improve training on non-IID data by creating a small subset of data which is globally shared between all the edge devices, and shows that accuracy can be increased by 30% for the CIFAR-10 dataset with only 5% globally shared data.
- OpenAlex ID:W2807006176
- OpenAlex DOI:10.48550/arxiv.1806.00582
- DOI:10.48550/arxiv.1806.00582
- DOI来源:OpenAlex
- 开放获取:green
- 开放获取链接:https://arxiv.org/pdf/1806.00582
- OpenAlex更新:2026-08-25
- 待LLM分类:否
- 成熟度:research
- 场景:federated learning、non-IID、heterogeneous data
- 标题中文:使用非独立同分布数据的联邦学习
- TLDR中文:本文提出了一种通过创建在所有边缘设备之间全局共享的小型数据子集来改进非独立同分布数据训练的策略,并表明在 CIFAR-10 数据集上,仅共享 5% 的全局数据即可将准确率提升 30%。
- 来源文件:
- [OpenAlex discover]
- [S2 enrich]