Summary of Federated Communication-efficient Multi-objective Optimization, by Baris Askin et al.
Federated Communication-Efficient Multi-Objective Optimization
by Baris Askin, Pranay Sharma, Gauri Joshi, Carlee Joe-Wong
First submitted to arxiv on: 21 Oct 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: 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 This paper presents FedCMOO, a novel federated multi-objective optimization (FMOO) algorithm that efficiently trains a single model to optimize multiple objective functions while minimizing communication costs. Unlike existing approaches, FedCMOO’s communication cost does not scale with the number of objectives, as each client sends a single aggregated gradient obtained using randomized singular value decomposition (SVD). The proposed method is analyzed for smooth non-convex objective functions under milder assumptions than prior work. A variant of FedCMOO allows users to specify a preference over the objectives in terms of a desired ratio of final objective values. Experimental results demonstrate the superiority of FedCMOO over baseline approaches. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper finds a way to make machines learn multiple things at once, like making something strong and efficient. It’s like training for a marathon and a bike ride at the same time! The new method, called FedCMOO, is really good because it only sends one message between the different parts of the machine, even if there are many goals to reach. This makes it faster and more efficient. The scientists who wrote this paper also made a special version that lets people say which goal is most important. It’s like saying “I want to be strong” or “I want to be fast”. This new way works really well and is better than the old ways. |
Keywords
* Artificial intelligence * Optimization
