Summary of Attribute Inference Attacks For Federated Regression Tasks, by Francesco Diana et al.
Attribute Inference Attacks for Federated Regression Tasks
by Francesco Diana, Othmane Marfoq, Chuan Xu, Giovanni Neglia, Frédéric Giroire, Eoin Thomas
First submitted to arxiv on: 19 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
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 The proposed study investigates the vulnerability of Federated Learning (FL) to reconstruction attacks, specifically attribute inference attacks (AIA), which can reveal sensitive attributes of targeted clients. The research focuses on regression tasks in FL environments and proposes novel model-based AIAs that consider scenarios where adversaries eavesdrop or directly interfere with the training process. The study benchmarks the proposed attacks against state-of-the-art methods using real-world datasets, demonstrating a significant increase in reconstruction accuracy, particularly in heterogeneous client datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated Learning helps devices like phones and smart home gadgets share data without sharing sensitive information. But attackers can use this shared data to figure out private details about individual devices. This is a problem because it can happen with more than just classification tasks – it can also affect regression tasks, which are used for things like predicting house prices or stock prices. The researchers came up with new ways for attackers to do this in FL and tested them on real-world datasets. They found that these attacks are pretty good at guessing private details about devices. |
Keywords
» Artificial intelligence » Classification » Federated learning » Inference » Regression
