Summary of A Stochastic Optimization Framework For Private and Fair Learning From Decentralized Data, by Devansh Gupta et al.
A Stochastic Optimization Framework for Private and Fair Learning From Decentralized Data
by Devansh Gupta, A.S. Poornash, Andrew Lowy, Meisam Razaviyayn
First submitted to arxiv on: 12 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
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 Machine learning models are trained on sensitive data distributed across different “silos” (e.g., hospitals). These federated learning models may then be used to make consequential decisions, such as allocating healthcare resources. The paper develops a novel algorithm for private and fair federated learning (FL) that satisfies inter-silo record-level differential privacy (ISRL-DP), promoting different fairness notions like demographic parity and equalized odds. The algorithm converges under mild smoothness assumptions on the loss function, whereas prior work required strong convexity for convergence. Experiments demonstrate the state-of-the-art fairness-accuracy tradeoffs of the algorithm across different privacy levels. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning models are trained on sensitive data from hospitals and other places. These models can be used to make important decisions like how to use healthcare resources. The problem is that these models need to keep personal data private, even if someone tries to figure out what’s in the data. Also, the decisions made by these models should be fair for all people, regardless of their race or gender. This paper creates a new way to do federated learning that keeps personal data private and makes sure the decisions are fair. |
Keywords
* Artificial intelligence * Federated learning * Loss function * Machine learning
