Robust and differentially private mean estimation

X Liu, W Kong, S Kakade, S Oh - Advances in neural …, 2021 - proceedings.neurips.cc
In statistical learning and analysis from shared data, which is increasingly widely adopted in
platforms such as federated learning and meta-learning, there are two major concerns …

Covariance-aware private mean estimation without private covariance estimation

G Brown, M Gaboardi, A Smith… - Advances in neural …, 2021 - proceedings.neurips.cc
We present two sample-efficient differentially private mean estimators for $ d $-dimensional
(sub) Gaussian distributions with unknown covariance. Informally, given $ n\gtrsim d/\alpha …

Choosing among notions of multivariate depth statistics

K Mosler, P Mozharovskyi - Statistical Science, 2022 - projecteuclid.org
Classical multivariate statistics measures the outlyingness of a point by its Mahalanobis
distance from the mean, which is based on the mean and the covariance matrix of the data …

A pseudo-metric between probability distributions based on depth-trimmed regions

G Staerman, P Mozharovskyi, P Colombo… - arXiv preprint arXiv …, 2021 - arxiv.org
The design of a metric between probability distributions is a longstanding problem motivated
by numerous applications in Machine Learning. Focusing on continuous probability …

Calibrated multiple-output quantile regression with representation learning

S Feldman, S Bates, Y Romano - Journal of Machine Learning Research, 2023 - jmlr.org
We develop a method to generate predictive regions that cover a multivariate response
variable with a user-specified probability. Our work is composed of two components. First …

Statistical depth functions for ranking distributions: definitions, statistical learning and applications

M Goibert, S Clémençon, E Irurozki… - arXiv preprint arXiv …, 2022 - arxiv.org
The concept of median/consensus has been widely investigated in order to provide a
statistical summary of ranking data, ie realizations of a random permutation $\Sigma $ of a …

Affine-invariant integrated rank-weighted depth: Definition, properties and finite sample analysis

G Staerman, P Mozharovskyi, S Clémençon - arXiv preprint arXiv …, 2021 - arxiv.org
Because it determines a center-outward ordering of observations in $\mathbb {R}^ d $ with $
d\geq 2$, the concept of statistical depth permits to define quantiles and ranks for …

Affine invariant integrated rank-weighted statistical depth: properties and finite sample analysis

S Clémençon, P Mozharovskyi… - Electronic Journal of …, 2023 - projecteuclid.org
Because it determines a center-outward ordering of observations in R d with d≥ 2, the
concept of statistical depth permits to define quantiles and ranks for multivariate data and …

Another look at halfspace depth: flag halfspaces with applications

D Pokorný, P Laketa, S Nagy - Journal of Nonparametric Statistics, 2024 - Taylor & Francis
The halfspace depth is a well-studied tool of nonparametric statistics in multivariate spaces.
We introduce a flag halfspace–an intermediary between a closed halfspace and its interior …

How to find a point in the convex hull privately

H Kaplan, M Sharir, U Stemmer - arXiv preprint arXiv:2003.13192, 2020 - arxiv.org
We study the question of how to compute a point in the convex hull of an input set $ S $ of $
n $ points in ${\mathbb R}^ d $ in a differentially private manner. This question, which is …