Summary of Treecss: An Efficient Framework For Vertical Federated Learning, by Qinbo Zhang et al.
TreeCSS: An Efficient Framework for Vertical Federated Learning
by Qinbo Zhang, Xiao Yan, Yukai Ding, Quanqing Xu, Chuang Hu, Xiaokai Zhou, Jiawei Jiang
First submitted to arxiv on: 3 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This research proposes an efficient vertical federated learning (VFL) framework called TreeCSS to accelerate the alignment and training processes in multi-party VFL scenarios. The framework consists of two main components: Tree-MPSI for sample alignment and coreset selection (CSS) for model training. Tree-MPSI adopts a tree-based structure and scheduling strategy to parallelize alignment among participants, while CSS uses clustering to select representative data samples for training. The framework is evaluated on various datasets and models, showing up to 2.93x acceleration in training time compared to vanilla VFL with comparable model accuracy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning lets different groups work together on a shared goal without sharing their individual data. This research makes it faster and more efficient for these groups to do so. The approach is called TreeCSS, and it has two main parts: one that helps align the data between groups (Tree-MPSI) and another that chooses the most important data samples for training (coreset selection). By using a tree-based structure and scheduling strategy, TreeCSS can process large amounts of data quickly. The results show that this approach is faster than traditional methods while still producing accurate models. |
Keywords
» Artificial intelligence » Alignment » Clustering » Federated learning
