Summary of Fliphat: Joint Differential Privacy For High Dimensional Sparse Linear Bandits, by Sunrit Chakraborty et al.
FLIPHAT: Joint Differential Privacy for High Dimensional Sparse Linear Bandits
by Sunrit Chakraborty, Saptarshi Roy, Debabrota Basu
First submitted to arxiv on: 22 May 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Machine Learning (cs.LG); Statistics Theory (math.ST)
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 proposes a framework for joint differentially private high dimensional sparse linear bandits, aiming to balance privacy concerns with optimal decision-making. In this setting, rewards and contexts are considered as private data, and the authors derive a lower bound on the regret achievable while ensuring privacy. To address the problem, they design a computationally efficient algorithm, FLIPHAT, which combines episodic forgetting, doubling of episodes, and sparse linear regression oracle to ensure both privacy and regret-optimality. The algorithm achieves optimal regret up to a linear factor in model sparsity and logarithmic factor in context dimension. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper explores how to make decisions while protecting user data in applications like personalized medicine. It’s about balancing what’s good for users with the need to keep their information private. The authors find a way to do this by designing an algorithm that makes decisions based on some information, but only uses a small part of it. This helps keep the rest of the information safe. They also show how well this algorithm works in terms of making good decisions while keeping data private. |
Keywords
» Artificial intelligence » Linear regression
