Summary of Fedac: An Adaptive Clustered Federated Learning Framework For Heterogeneous Data, by Yuxin Zhang et al.
FedAC: An Adaptive Clustered Federated Learning Framework for Heterogeneous Data
by Yuxin Zhang, Haoyu Chen, Zheng Lin, Zhe Chen, Jin Zhao
First submitted to arxiv on: 25 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
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 In this paper, researchers propose an adaptive clustered federated learning (CFL) framework called FedAC to address the limitations of current CFL methods. These limitations include inadequate integration of global and intra-cluster knowledge and a lack of an efficient online model similarity metric. The proposed framework consists of three main components: (1) a neural network architecture that decouples submodules and utilizes distinct aggregation methods for each, allowing for more effective integration of global knowledge; (2) a cost-effective online model similarity metric based on dimensionality reduction to identify similar models; and (3) a cluster number fine-tuning module for improved adaptability and scalability. The authors evaluate FedAC using extensive experiments on CIFAR-10 and CIFAR-100 datasets, achieving superior empirical performance compared to state-of-the-art methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research paper presents a new way of learning called adaptive clustered federated learning (CFL). CFL helps machines learn from many different sources of data. The authors found that current CFL methods don’t work well because they don’t use all the available information and can’t decide which groups to put similar data into. To solve this problem, they created a new framework called FedAC. It has three parts: one that combines information from different groups, another that compares models to find similar ones, and a third that adjusts the group sizes as needed. The authors tested their method on two big datasets and found it worked better than previous methods. |
Keywords
* Artificial intelligence * Dimensionality reduction * Federated learning * Fine tuning * Neural network
