Summary of Map: Model Aggregation and Personalization in Federated Learning with Incomplete Classes, by Xin-chun Li et al.
MAP: Model Aggregation and Personalization in Federated Learning with Incomplete Classes
by Xin-Chun Li, Shaoming Song, Yinchuan Li, Bingshuai Li, Yunfeng Shao, Yang Yang, De-Chuan Zhan
First submitted to arxiv on: 14 Apr 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 proposes a new approach to federated learning (FL) that addresses the challenges posed by non-independent and identically distributed (Non-I.I.D.) data samples on local devices. In particular, it focuses on scenarios where clients own incomplete classes, i.e., each client can only access a partial set of the whole class set. To tackle this issue, the authors introduce “restricted softmax” as an alternative to standard softmax for model aggregation and propose “inherited private model” for better model personalization. The resulting algorithm, MAP, simultaneously achieves the aggregation and personalization goals in FL. Experimental studies demonstrate the superiority of MAP over existing methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about a way to make machines learn together without sharing sensitive information. Imagine you’re trying to teach a robot to recognize different animals, but each robot only has pictures of some of those animals. The challenge is how to combine all that knowledge into one good animal-recognizing model. To solve this problem, the researchers suggest new ways to do two things: first, combining the models from each robot without sharing their data; and second, helping each robot improve its own recognition skills based on what it’s learned so far. They call this approach MAP, and they show through experiments that it works better than other methods. |
Keywords
» Artificial intelligence » Federated learning » Softmax
