Abstract: We introduce a family of multiscale stick-breaking mixture models for Bayesian nonparametric density estimation. The Bayesian nonparametric literature is dominated by single scale methods, exception made for P`olya trees and allied approaches. Our proposal is based on a mixture specification exploiting an infinitely-deep binary tree of random weights that grows according to a multiscale generalization of a large class of stick-breaking processes; this multiscale stick-breaking is paired with specific stochastic processes generating sequences of parameters that induce stochastically ordered kernel functions. Properties of this family of multiscale stick-breaking mixtures are described. Focusing on a Gaussian specification, a Markov Chain Montecarlo algorithm for posterior computation is introduced. The performance of the method is illustrated analyzing both synthetic and real data sets. The method is well-suited for data living in and is able to detect densities with varying degree of smoothness and local features.
Recommendations
- SCALE MIXTURES DISTRIBUTIONS IN STATISTICAL MODELLING
- Mixtures of marginal models
- Independent Component Analysis and Blind Signal Separation
- Finite mixtures of structured models
- Decomposition of models whose marginal distributions are mixtures
- Multivariate stochastic comparisons of mixture models
- Finite mixture models
- Mixture models and applications
Cites work
- A Bayesian analysis of some nonparametric problems
- Bayesian Density Estimation and Inference Using Mixtures
- Bayesian density estimation using bernstein polynomials
- Bayesian inference with dependent normalized completely random measures
- Bayesian Kernel Mixtures for Counts
- Computationally tractable approximate and smoothed Polya trees
- Conjugacy as a Distinctive Feature of the Dirichlet Process
- Default priors for density estimation with mixture models
- Density Estimation With Confidence Sets Exemplified by Superclusters and Voids in the Galaxies
- Distributional results for means of normalized random measures with independent increments
- Exact mean integrated squared error
- Gibbs Sampling Methods for Stick-Breaking Priors
- scientific article; zbMATH DE number 597911 (Why is no real title available?)
- scientific article; zbMATH DE number 774881 (Why is no real title available?)
- More aspects of Pólya tree distributions for statistical modelling
- Multiscale Bernstein polynomials for densities
- Normalized random measures driven by increasing additive processes
- On a class of Bayesian nonparametric estimates: I. Density estimates
- Optional Pólya tree and Bayesian inference
- Polya trees and random distributions
- Posterior Analysis for Normalized Random Measures with Independent Increments
- Random Bernstein Polynomials
- Robustifying Bayesian nonparametric mixtures for count data
- Shared kernel Bayesian screening
- Size-biased sampling of Poisson point processes and excursions
- Slice sampling mixture models
- Some aspects of Pólya tree distributions for statistical modelling
- The two-parameter Poisson-Dirichlet distribution derived from a stable subordinator
Cited in
(8)- Bayesian multivariate mixed-scale density estimation
- Mixture modeling on related samples by \(\psi\)-stick breaking and kernel perturbation
- Some further developments for stick-breaking priors: finite and infinite clustering and classifica\-tion
- An extended mixture model for the simultaneous treatment of small‐scale and large‐scale interfaces
- Multiscale Bernstein polynomials for densities
- Mixture polarization in inter-rater agreement analysis: a Bayesian nonparametric index
- A tree perspective on stick-breaking models in covariate-dependent mixtures (with discussion)
- Markov stick-breaking processes
This page was built for publication: Multiscale stick-breaking mixture models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2029098)
