Uncertainty quantification for Bayesian CART
From MaRDI portal
Publication:2073718
DOI10.1214/21-AOS2093zbMATH Open1486.62131arXiv1910.07635OpenAlexW4205545960MaRDI QIDQ2073718FDOQ2073718
Authors: Ismaël Castillo, Veronika Rockova 
Publication date: 7 February 2022
Published in: The Annals of Statistics (Search for Journal in Brave)
Abstract: This work affords new insights into Bayesian CART in the context of structured wavelet shrinkage. The main thrust is to develop a formal inferential framework for Bayesian tree-based regression. We reframe Bayesian CART as a g-type prior which departs from the typical wavelet product priors by harnessing correlation induced by the tree topology. The practically used Bayesian CART priors are shown to attain adaptive near rate-minimax posterior concentration in the supremum norm in regression models. For the fundamental goal of uncertainty quantification, we construct adaptive confidence bands for the regression function with uniform coverage under self-similarity. In addition, we show that tree-posteriors enable optimal inference in the form of efficient confidence sets for smooth functionals of the regression function.
Full work available at URL: https://arxiv.org/abs/1910.07635
Recommendations
posterior concentrationrecursive partitioningnonparametric Bernstein-von Mises theoremregression treesBayesian CART
Asymptotic properties of nonparametric inference (62G20) Nonparametric tolerance and confidence regions (62G15)
Cites Work
- Variable selection for BART: an application to gene regulation
- BART: Bayesian additive regression trees
- Consistency of random forests
- Title not available (Why is that?)
- A Bayesian CART algorithm
- Estimation and Inference of Heterogeneous Treatment Effects using Random Forests
- Optimal predictive model selection.
- Quantifying uncertainty in random forests via confidence intervals and hypothesis tests
- Mathematical foundations of infinite-dimensional statistical models
- Title not available (Why is that?)
- Fundamentals of nonparametric Bayesian inference
- Convergence rates of posterior distributions.
- The risk inflation criterion for multiple regression
- Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects (with discussion)
- Wavelets on the interval and fast wavelet transforms
- Pólya tree posterior distributions on densities
- Nonparametric statistical inference for drift vector fields of multi-dimensional diffusions
- Bernstein-von Mises theorems for statistical inverse problems. II: Compound Poisson processes
- Nonparametric Bernstein-von Mises theorems in Gaussian white noise
- On the Bernstein-von Mises phenomenon for nonparametric Bayes procedures
- On Bayesian supremum norm contraction rates
- A Reference Bayesian Test for Nested Hypotheses and its Relationship to the Schwarz Criterion
- Benchmark priors for Bayesian model averaging.
- Optional Pólya tree and Bayesian inference
- Asymptotic equivalence of nonparametric regression and white noise
- Model Selection for CART Regression Trees
- Criteria for Bayesian model choice with application to variable selection
- Lower bounds for posterior rates with Gaussian process priors
- Frequentist coverage of adaptive nonparametric Bayesian credible sets
- Bayesian optimal adaptive estimation using a sieve prior
- Analysis of a random forests model
- On adaptive posterior concentration rates
- Rates of contraction for posterior distributions in \(L^{r}\)-metrics, \(1 \leq r \leq \infty\)
- Supremum norm posterior contraction and credible sets for nonparametric multivariate regression
- Adaptive Bayesian density estimation in \(L^p\)-metrics with Pitman-Yor or normalized inverse-Gaussian process kernel mixtures
- Model-Based Compressive Sensing
- Honest adaptive confidence bands and self-similar functions
- Learning Theory
- Adaptive Bayesian Wavelet Shrinkage
- CART and best-ortho-basis: a connection
- Minimax-optimal classification with dyadic decision trees
- A simple wavelet approach to nonparametric regression from recursive partitioning schemes
- A new class of unbalanced Haar wavelets that form an unconditional basis for \(L_ p\) on general measure spaces
- Unbalanced Haar Technique for Nonparametric Function Estimation
- Posterior concentration for Bayesian regression trees and forests
- Adaptive Bernstein-von Mises theorems in Gaussian white noise
- Multiscale Poisson Intensity and Density Estimation
- On information pooling, adaptability and superefficiency in nonparametric function estimation
- Bayesian regression trees for high-dimensional prediction and variable selection
- Bayesian regression tree ensembles that adapt to smoothness and sparsity
- Spike and slab Pólya tree posterior densities: adaptive inference
Cited In (5)
- Smoothing and adaptation of shifted Pólya tree ensembles
- Posterior concentration for Bayesian regression trees and forests
- Heavy-tailed Bayesian nonparametric adaptation
- Optional Pólya trees: posterior rates and uncertainty quantification
- Multidimensional adaptive relevance vector machines for uncertainty quantification
Uses Software
This page was built for publication: Uncertainty quantification for Bayesian CART
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2073718)
