Robust and scalable Bayes via a median of subset posterior measures
From MaRDI portal
(Redirected from Publication:4637048)
Abstract: We propose a novel approach to Bayesian analysis that is provably robust to outliers in the data and often has computational advantages over standard methods. Our technique is based on splitting the data into non-overlapping subgroups, evaluating the posterior distribution given each independent subgroup, and then combining the resulting measures. The main novelty of our approach is the proposed aggregation step, which is based on the evaluation of a median in the space of probability measures equipped with a suitable collection of distances that can be quickly and efficiently evaluated in practice. We present both theoretical and numerical evidence illustrating the improvements achieved by our method.
Recommendations
Cites work
- A Bayesian approach to some outlier problems
- A quadratically convergent method for minimizing a sum of euclidean norms
- A Scalable Bootstrap for Massive Data
- Algebraic optimization: The Fermat-Weber location problem
- An overview of robust Bayesian analysis. (With discussion)
- Bayesian model robustness via disparities
- Breakdown points of affine equivariant estimators of multivariate location and covariance matrices
- Consistent and robust Bayes procedures for location based on partial information
- Convergence rates of posterior distributions.
- Distributed optimization and statistical learning via the alternating direction method of multipliers
- Efficient and fast estimation of the geometric median in Hilbert spaces with an averaged stochastic gradient algorithm
- Extending the rank likelihood for semiparametric copula estimation
- Fast approximations for sums of distances, clustering and the Fermat-Weber problem
- Gaussian processes for machine learning.
- Geometric median and robust estimation in Banach spaces
- Hilbert space embeddings and metrics on probability measures
- scientific article; zbMATH DE number 6377992 (Why is no real title available?)
- scientific article; zbMATH DE number 3790208 (Why is no real title available?)
- scientific article; zbMATH DE number 42272 (Why is no real title available?)
- scientific article; zbMATH DE number 1256715 (Why is no real title available?)
- scientific article; zbMATH DE number 1827860 (Why is no real title available?)
- scientific article; zbMATH DE number 863578 (Why is no real title available?)
- scientific article; zbMATH DE number 3025040 (Why is no real title available?)
- Loss minimization and parameter estimation with heavy tails
- Nonparametric Bayes modeling of multivariate categorical data
- On the Convergence of a Class of Iterative Methods for Solving the Weber Location Problem
- On the empirical estimation of integral probability metrics
- On the mean speed of convergence of empirical and occupation measures in Wasserstein distance
- On the rate of convergence in Wasserstein distance of the empirical measure
- Pattern recognition and machine learning.
- PRMLT
- Probability inequalities for likelihood ratios and convergence rates of sieve MLEs
- Recursive estimation of the conditional geometric median in Hilbert spaces
- Robust Bayesian inference via coarsening
- Robust Statistics
- Scalable Bayes via barycenter in Wasserstein space
- Weak convergence and empirical processes. With applications to statistics
- Weiszfeld's method: old and new results
Cited in
(36)- Robust and parallel Bayesian model selection
- Joining and splitting models with Markov melding
- Bayesian bootstraps for massive data
- A case study competition among methods for analyzing large spatial data
- Distributed statistical estimation and rates of convergence in normal approximation
- Distributed Bayesian matrix factorization with limited communication
- Double-parallel Monte Carlo for Bayesian analysis of big data
- Accelerated dimension-independent adaptive metropolis
- Semidual regularized optimal transport
- Scalable Bayes via barycenter in Wasserstein space
- scientific article; zbMATH DE number 7008320 (Why is no real title available?)
- Distributed Bayesian Inference in Linear Mixed-Effects Models
- An asymptotic analysis of distributed nonparametric methods
- Hamiltonian Monte Carlo with energy conserving subsampling
- Nonparametric Bayesian aggregation for massive data
- scientific article; zbMATH DE number 7164769 (Why is no real title available?)
- Stochastic gradient Markov chain Monte Carlo
- Bayesian computation: a summary of the current state, and samples backwards and forwards
- Randomized maximum-contrast selection: subagging for large-scale regression
- Linear-Cost Covariance Functions for Gaussian Random Fields
- Distributed Bayesian inference in massive spatial data
- Divide-and-conquer Metropolis-Hastings samplers with matched samples
- Divide-and-conquer Bayesian inference in hidden Markov models
- An algorithm for distributed Bayesian inference
- Bayesian Robustness: A Nonasymptotic Viewpoint
- Insufficient Gibbs sampling
- Robust optimization and inference on manifolds
- A Scalable Partitioned Approach to Model Massive Nonstationary Non-Gaussian Spatial Datasets
- Frequentist guarantees of distributed (non)-Bayesian inference
- Differentially Private Methods for Compositional Data
- Selective inference with distributed data
- Optimizing large-scale educational assessment with a ``divide-and-conquer strategy: fast and efficient distributed Bayesian inference in IRT models
- Robust probabilistic inference via a constrained transport metric (with discussion)
- Clustering the nearest neighbor Gaussian process
- Scalable Gaussian process regression via median posterior inference for estimating the health effects of an environmental mixture
- Bayesian mixture modeling for multivariate conditional distributions
This page was built for publication: Robust and scalable Bayes via a median of subset posterior measures
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4637048)
