Hybrid Wasserstein distance and fast distribution clustering
From MaRDI portal
(Redirected from Publication:2283573)
Abstract: We define a modified Wasserstein distance for distribution clustering which inherits many of the properties of the Wasserstein distance but which can be estimated easily and computed quickly. The modified distance is the sum of two terms. The first term --- which has a closed form --- measures the location-scale differences between the distributions. The second term is an approximation that measures the remaining distance after accounting for location-scale differences. We consider several forms of approximation with our main emphasis being a tangent space approximation that can be estimated using nonparametric regression. We evaluate the strengths and weaknesses of this approach on simulated and real examples.
The Wasserstein distance became one of the most popular and promising tools in machine learning community lately because of its high ability to capture the underlying geometry, possibility to compare the discrete and continuous distributions and shape preservation. The main contribution of this paper is the efficient modification of the Wasserstein distance that can be estimated and computed quickly. The efficiency of a proposed hybrid Wasserstein \(k\)-means clustering methods is illustrated on univariate and multivariate examples.
Recommendations
- Robust clustering tools based on optimal transportation
- Wasserstein distance on finite spaces: statistical inference and algorithms
- Inference for empirical Wasserstein distances on finite spaces
- Fuzzy clustering of distributional data with automatic weighting of variable components
- Clustering, factor discovery and optimal transport
Cites work
- A class of Wasserstein metrics for probability distributions
- A fixed-point approach to barycenters in Wasserstein space
- A linear optimal transportation framework for quantifying and visualizing variations in sets of images
- A population background for nonparametric density-based clustering
- An Exact Distribution-Free Test Comparing Two Multivariate Distributions based on Adjacency
- Data-driven distributionally robust optimization using the Wasserstein metric: performance guarantees and tractable reformulations
- Energy statistics: a class of statistics based on distances
- Fréchet means and Procrustes analysis in Wasserstein space
- Functional mixed effects models
- scientific article; zbMATH DE number 6381735 (Why is no real title available?)
- scientific article; zbMATH DE number 1909499 (Why is no real title available?)
- scientific article; zbMATH DE number 7415088 (Why is no real title available?)
- scientific article; zbMATH DE number 3231692 (Why is no real title available?)
- Inference for empirical Wasserstein distances on finite spaces
- Limit laws of the empirical Wasserstein distance: Gaussian distributions
- Multivariate kernel smoothing and its applications
- Nonparametric functional data analysis. Theory and practice.
- On the Bures-Wasserstein distance between positive definite matrices
- Optimal transport: fast probabilistic approximation with exact solvers
- Preconditioning of optimal transport
- Robust clustering tools based on optimal transportation
- Sharp asymptotic and finite-sample rates of convergence of empirical measures in Wasserstein distance
Cited in
(19)- Wasserstein discriminant analysis
- Hierarchical clustering with optimal transport
- Minimax confidence intervals for the sliced Wasserstein distance
- Linear regression for numeric symbolic variables: a least squares approach based on Wasserstein distance
- Co-clustering algorithms for distributional data with automated variable weighting
- Network consensus in the Wasserstein metric space of probability measures
- On clustering uncertain and structured data with Wasserstein barycenters and a geodesic criterion for the number of clusters
- Wasserstein distance on finite spaces: statistical inference and algorithms
- Fast Discrete Distribution Clustering Using Wasserstein Barycenter With Sparse Support
- Clustering, factor discovery and optimal transport
- scientific article; zbMATH DE number 7415088 (Why is no real title available?)
- Multisource Single-Cell Data Integration by MAW Barycenter for Gaussian Mixture Models
- On the use of Wasserstein distance in the distributional analysis of human decision making under uncertainty
- Central limit theorems for semi-discrete Wasserstein distances
- Covariance-based soft clustering of functional data based on the Wasserstein-Procrustes metric
- A note on the radiant formula and its relations to the sliced Wasserstein distance
- Asymptotic theory of geometric and adaptive k-means clustering
- Wasserstein k-centers clustering for distributional data
- Bayesian Wasserstein repulsive Gaussian mixture models
This page was built for publication: Hybrid Wasserstein distance and fast distribution clustering
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2283573)
