Summary of Noisy Data Meets Privacy: Training Local Models with Post-processed Remote Queries, by Kexin Li et al.
Noisy Data Meets Privacy: Training Local Models with Post-Processed Remote Queries
by Kexin Li, Aastha Mehta, David Lie
First submitted to arxiv on: 25 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Cryptography and Security (cs.CR); Computers and Society (cs.CY)
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 paper LDPKiT addresses concerns about end-user data privacy in cloud-based models used for inference in privacy-sensitive domains. It proposes a two-layer noise injection framework that leverages locally differentially private (LDP) and its post-processing property to create a privacy-protected inference dataset. This dataset is then used to train a reliable local model for sensitive inputs. The paper demonstrates that LDPKiT effectively improves utility while preserving privacy on datasets such as Fashion-MNIST, SVHN, and PathMNIST medical datasets. The results show that at higher noise levels, LDPKiT achieves similar inference accuracy with stronger privacy guarantees. Additionally, the authors perform sensitivity analyses to evaluate the impact of dataset sizes and analyze latent space representations to explain the accuracy improvements. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps solve a problem where big models on the cloud can’t keep our personal data private. They created a new way to make sure our data stays safe while still getting accurate results. This method, called LDPKiT, works by adding special noise to the data before sending it to the cloud. The paper shows that this approach keeps our data private and still gives good results on different types of datasets. |
Keywords
» Artificial intelligence » Inference » Latent space
