arXiv research

Improved private geometric median estimation with nearly-linear time complexity.

New method for estimating median and mean with high probability privacy.

DiPriMe forests use private medians to create balanced tree splits for privacy-protected data.

Develops privacy-preserving multivariate median estimation methods.

Paper addresses robust federated linear bandits against Byzantine attacks.

This paper proposes methods to compute differentially private confidence intervals for the median.

Private statistical inference methods improve confidence interval lengths.

Efficiently private clustering algorithms with tight approximation ratios.

We present a private learner for halfspaces over an arbitrary finite domain $X\subset \mathbb{R}^d$ with sample complexity $mathrm{poly}(d,2^{\log^*|X|})$. The building block for this learner is a differentially private algorithm for locating an approximate center point of $m>\mathrm{poly}(d,2^{\log^*|X|})$ points -- a…

We tackle the problem of estimating a location parameter with differential privacy guarantees and sub-Gaussian deviations. Recent work in statistics has focused on the study of estimators that achieve sub-Gaussian type deviations even for heavy tailed data. We revisit some of these estimators through the lens of differ…

This paper introduces online algorithms to estimate robust geometric median in large data streams.

In high dimensions, the mean and geometric median are nearly identical.

New method for better initial centers in clustering with improved accuracy and privacy.

Paper proposes a robust method for federated ICA with geometric median aggregation.

Paper shows how to use geometric median for robust SGD in high dimensions.

Combines public and private data for better statistical estimation.

We show that uniform lattices of isometries of products of real hyperbolic spaces act properly discontinuously and cocompactly on a median space. For lattices in products of at least two factors, this is the strongest degree of compatibility possible with the median geometry. Our theorem is also relevant for potential …

We design a new algorithm for the Euclidean $k$-means problem that operates in the local model of differential privacy. Unlike in the non-private literature, differentially private algorithms for the $k$-means objective incur both additive and multiplicative errors. Our algorithm significantly reduces the additive erro…

We develop a coreset for robust geometric median, reducing size dependency on outliers.

In this paper, we define the geometric median of a probability measure on a Riemannian manifold, give its characterization and a natural condition to ensure its uniqueness. In order to calculate the median in practical cases, we also propose a subgradient algorithm and prove its convergence as well as estimating the er…

Gradient clipping helps private SGD converge despite potential bias.

Novel algorithm reduces privacy noise in machine learning.

Boosting framework for vector-valued prediction with geometric stability.

In this paper, we initiate a systematic investigation of differentially private algorithms for convex empirical risk minimization. Various instantiations of this problem have been studied before. We provide new algorithms and matching lower bounds for private ERM assuming only that each data point's contribution to the…

This work achieves exponential concentration in heavy-tailed data over CAT(κ) spaces using the Fréchet median.

Federated learning has a variety of applications in multiple domains by utilizing private training data stored on different devices. However, the aggregation process in federated learning is highly vulnerable to adversarial attacks so that the global model may behave abnormally under attacks. To tackle this challenge, …

This study presents a long-term alternative formula for stock price variation described by a geometric Brownian motion on the basis of median instead of mean or expected values. The proposed method is motivated by the observation made in remote fields, where optimality of bet-hedging or diversification strategies is ex…

Byrd-SAGA reduces variance to robustify SGD against Byzantine attacks.

We construct compactifications for median spaces with compact intervals, generalising Roller boundaries of ${\rm CAT}(0)$ cube complexes. Examples of median spaces with compact intervals include all finite rank median spaces and all proper median spaces of infinite rank. Our methods also work for general median algebra…

Minimizing a convex risk function is the main step in many basic learning algorithms. We study protocols for convex optimization which provably leak very little about the individual data points that constitute the loss function. Specifically, we consider differentially private algorithms that operate in the local model…

MCE reduces embedding instability in nonlinear dimensionality reduction.

Accelerated optimization methods improve robustness and privacy in estimation.

New concept of coarse medians for higher rank symmetric spaces.

Improves clustering interpretability with decision trees.

Unique median structures found in hyperbolic spaces.

Study on median algebra structures on Euclidean spaces and manifolds with local CAT(0) cubulation.

Robust score matching improves parameter estimation in contaminated data.

We prove a version of the Tits alternative for groups acting on complete, finite rank median spaces. This shows that group actions on finite rank median spaces are much more restricted than actions on general median spaces. Along the way, we extend to median spaces the Caprace-Sageev machinery and part of Hagen's theor…

Distributed model training is vulnerable to byzantine system failures and adversarial compute nodes, i.e., nodes that use malicious updates to corrupt the global model stored at a parameter server (PS). To guarantee some form of robustness, recent work suggests using variants of the geometric median as an aggregation r…

Convex cores found for group actions on median spaces.

We introduce and begin to explore the mean and median of finite sets of shapes represented as integral currents. The median can be computed efficiently in practice, and we focus most of our theoretical and computational attention on medians. We consider questions on the existence and regularity of medians. While the me…

We present a general method for privacy-preserving Bayesian inference in Poisson factorization, a broad class of models that includes some of the most widely used models in the social sciences. Our method satisfies limited precision local privacy, a generalization of local differential privacy, which we introduce to fo…

Positive definite matrices abound in a dazzling variety of applications. This ubiquity can be in part attributed to their rich geometric structure: positive definite matrices form a self-dual convex cone whose strict interior is a Riemannian manifold. The manifold view is endowed with a "natural" distance function whil…

Novel algorithm resists Byzantine attacks in federated learning for PCA and LRCS.

We present differentially private efficient algorithms for learning union of polygons in the plane (which are not necessarily convex). Our algorithms achieve $(α,β)$-PAC learning and $(ε,δ)$-differential privacy using a sample of size $\tilde{O}\left(\frac{1}{αε}k\log d\right)$, where the domain is $[d]\times[d]$ and $…

Data is continuously generated by modern data sources, and a recent challenge in machine learning has been to develop techniques that perform well in an incremental (streaming) setting. In this paper, we investigate the problem of private machine learning, where as common in practice, the data is not given at once, but…

This paper is a short summary of our recent work on the medians and means of probability measures in Riemannian manifolds. Firstly, the existence and uniqueness results of local medians are given. In order to compute medians in practical cases, we propose a subgradient algorithm and prove its convergence. After that, F…

Datasets are often reused to perform multiple statistical analyses in an adaptive way, in which each analysis may depend on the outcomes of previous analyses on the same dataset. Standard statistical guarantees do not account for these dependencies and little is known about how to provably avoid overfitting and false d…