Summary of Fedadmm-insa: An Inexact and Self-adaptive Admm For Federated Learning, by Yongcun Song et al.
FedADMM-InSa: An Inexact and Self-Adaptive ADMM for Federated Learning
by Yongcun Song, Ziqi Wang, Enrique Zuazua
First submitted to arxiv on: 21 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR); Distributed, Parallel, and Cluster Computing (cs.DC); Optimization and Control (math.OC)
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 innovative federated learning (FL) algorithm called FedADMM-InSa that addresses challenges in developing efficient FL methods. The existing FedADMM algorithms show promise in handling heterogeneous data and systems but suffer from performance degradation if hyperparameters are not carefully tuned. To overcome this issue, the proposed algorithm eliminates the need for empirically setting local training accuracy by introducing an inexactness criterion. This allows each client to independently assess its unique condition, reducing local computational cost and mitigating straggle effects. The paper also presents a self-adaptive scheme that dynamically adjusts penalty parameters for each client, enhancing robustness. Numerical experiments on synthetic and real-world datasets demonstrate the proposed algorithm’s effectiveness in reducing clients’ local computational load and accelerating the learning process. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning is a way to learn from lots of data without sharing it all. This helps keep personal information private. The paper proposes a new method for doing this, called FedADMM-InSa. It makes an old algorithm better by letting each device decide how much work to do instead of following rules. This helps the devices use less energy and get results faster. The paper shows that this new method works well on some examples. |
Keywords
* Artificial intelligence * Federated learning
