理解私有进化作为学习增强的聚类
Understanding Private Evolution as Learning-Augmented Clustering
浏览论文内容
中文总结 AI 辅助
本文重新诠释私有进化算法为生成模型增强的Wasserstein学习，理论证明其性能界限可依赖内在维度，并提出几何感知版本以解决标准变体在良好聚类实例上的收敛问题，实验表明新算法具有竞争力且能提高召回率。
中文摘要 AI 辅助
私有进化（PE）是一种用于合成数据生成的差分隐私算法。虽然它可以被视为一种Wasserstein学习算法，但在实践中其表现远优于最坏情况下的Wasserstein分析所预测的结果。我们将PE重新解释为生成模型增强的Wasserstein学习。我们从理论上证明，当考虑到使用能够捕捉真实分布某些信息的生成模型时，可以获得更好的性能界限。例如，如果生成器提供的样本与分布处于相同的低维空间，则样本复杂度取决于内在维度而非环境维度。我们还表明，标准变体的PE在简单的良好聚类实例上可能无法收敛，并提出了一个新的几何感知版本的PE，该版本在此类实例上具有可证明的收敛性。实验上，我们展示了新算法与标准基线相比具有竞争力，并能提高召回率。
英文摘要
Private Evolution (PE) is a differentially private algorithm for synthetic data generation. While it can be viewed as a Wasserstein learning algorithm, it performs much better in practice than worst-case Wasserstein analyses would predict. We recast PE as generative model-augmented Wasserstein learning. We show theoretically that when we take into account the use of a generative model that is able to capture something about the true distribution, then we can obtain much better performance bounds. For example, if the generator gives samples in the same low-dimensional space as the distribution, then sample complexity depends on intrinsic, not ambient, dimension. We also show that standard variants of PE can fail to converge on simple well-clustered instances, and propose a new geometry-aware version of PE with provable convergence on such instances. Experimentally, we show that our new algorithm is competitive with standard baselines and can improve recall.
发表机构
- Apple（苹果公司）
- Yale University（耶鲁大学）
机构由 AI 辅助整理，请以论文原文为准。
