Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ashtiani, Hassan, Majid, Mahbod, Narayanan, Shyam
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2411.02298
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866915004563849216
author Ashtiani, Hassan
Majid, Mahbod
Narayanan, Shyam
author_facet Ashtiani, Hassan
Majid, Mahbod
Narayanan, Shyam
contents We study the problem of learning mixtures of Gaussians with approximate differential privacy. We prove that roughly $kd^2 + k^{1.5} d^{1.75} + k^2 d$ samples suffice to learn a mixture of $k$ arbitrary $d$-dimensional Gaussians up to low total variation distance, with differential privacy. Our work improves over the previous best result [AAL24b] (which required roughly $k^2 d^4$ samples) and is provably optimal when $d$ is much larger than $k^2$. Moreover, we give the first optimal bound for privately learning mixtures of $k$ univariate (i.e., $1$-dimensional) Gaussians. Importantly, we show that the sample complexity for privately learning mixtures of univariate Gaussians is linear in the number of components $k$, whereas the previous best sample complexity [AAL21] was quadratic in $k$. Our algorithms utilize various techniques, including the inverse sensitivity mechanism [AD20b, AD20a, HKMN23], sample compression for distributions [ABDH+20], and methods for bounding volumes of sumsets.
format Preprint
id arxiv_https___arxiv_org_abs_2411_02298
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Sample-Efficient Private Learning of Mixtures of Gaussians
Ashtiani, Hassan
Majid, Mahbod
Narayanan, Shyam
Machine Learning
Data Structures and Algorithms
Statistics Theory
We study the problem of learning mixtures of Gaussians with approximate differential privacy. We prove that roughly $kd^2 + k^{1.5} d^{1.75} + k^2 d$ samples suffice to learn a mixture of $k$ arbitrary $d$-dimensional Gaussians up to low total variation distance, with differential privacy. Our work improves over the previous best result [AAL24b] (which required roughly $k^2 d^4$ samples) and is provably optimal when $d$ is much larger than $k^2$. Moreover, we give the first optimal bound for privately learning mixtures of $k$ univariate (i.e., $1$-dimensional) Gaussians. Importantly, we show that the sample complexity for privately learning mixtures of univariate Gaussians is linear in the number of components $k$, whereas the previous best sample complexity [AAL21] was quadratic in $k$. Our algorithms utilize various techniques, including the inverse sensitivity mechanism [AD20b, AD20a, HKMN23], sample compression for distributions [ABDH+20], and methods for bounding volumes of sumsets.
title Sample-Efficient Private Learning of Mixtures of Gaussians
topic Machine Learning
Data Structures and Algorithms
Statistics Theory
url https://arxiv.org/abs/2411.02298