Essec\Faculty\Model\Contribution {#2216 ▼
#_index: "academ_contributions"
#_id: "12750"
#_source: array:26 [
"id" => "12750"
"slug" => "12750-coupling-based-convergence-assessment-of-some-gibbs-samplers-for-high-dimensional-bayesian-regression-with-shrinkage-priors 
12750-coupling-based-convergence-assessment-of-some-gibbs-samplers-for-high-dimensional-bayesian-reg "
"yearMonth" => "2022-07"
"year" => "2022"
"title" => "Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regression with shrinkage priors Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regressio "
"description" => "BISWAS, N., BHATTACHARYA, A., JACOB, P. et JOHNDROW, J. (2022). Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regression with shrinkage priors. <i>Journal of the Royal Statistical Society: Series B (Statistical Methodology)</i>, 84, pp. 973-996. BISWAS, N., BHATTACHARYA, A., JACOB, P. et JOHNDROW, J. (2022). Coupling-based convergence assessmen "
"authors" => array:4 [
0 => array:3 [
"name" => "JACOB Pierre"
"bid" => "B00795650"
"slug" => "jacob-pierre"
]
1 => array:1 [
"name" => "BISWAS Niloy"
]
2 => array:1 [
"name" => "BHATTACHARYA Anirban"
]
3 => array:1 [
"name" => "JOHNDROW James"
]
]
"ouvrage" => ""
"keywords" => array:5 [
0 => "Bayesian inference"
1 => "couplings"
2 => "Gibbs sampling"
3 => "Horseshoe prior"
4 => "parallel computation"
]
"updatedAt" => "2023-08-25 08:53:02"
"publicationUrl" => "https://doi.org/10.1111/rssb.12495"
"publicationInfo" => array:3 [
"pages" => "973-996"
"volume" => "84"
"number" => ""
]
"type" => array:2 [
"fr" => "Articles"
"en" => "Journal articles"
]
"support_type" => array:2 [
"fr" => "Revue scientifique"
"en" => "Scientific journal"
]
"countries" => array:2 [
"fr" => null
"en" => null
]
"abstract" => array:2 [
"fr" => "We consider Markov chain Monte Carlo (MCMC) algorithms for Bayesian high-dimensional regression with continuous shrinkage priors. A common challenge with these algorithms is the choice of the number of iterations to perform. This is critical when each iteration is expensive, as is the case when dealing with modern data sets, such as genome-wide association studies with thousands of rows and up to hundreds of thousands of columns. We develop coupling techniques tailored to the setting of high-dimensional regression with shrinkage priors, which enable practical, non-asymptotic diagnostics of convergence without relying on traceplots or long-run asymptotics. By establishing geometric drift and minorization conditions for the algorithm under consideration, we prove that the proposed couplings have finite expected meeting time. Focusing on a class of shrinkage priors which includes the ‘Horseshoe’, we empirically demonstrate the scalability of the proposed couplings. A highlight of our findings is that less than 1000 iterations can be enough for a Gibbs sampler to reach stationarity in a regression on 100,000 covariates. The numerical results also illustrate the impact of the prior on the computational efficiency of the coupling, and suggest the use of priors where the local precisions are Half-t distributed with degree of freedom larger than one. We consider Markov chain Monte Carlo (MCMC) algorithms for Bayesian high-dimensional regression with "
"en" => "We consider Markov chain Monte Carlo (MCMC) algorithms for Bayesian high-dimensional regression with continuous shrinkage priors. A common challenge with these algorithms is the choice of the number of iterations to perform. This is critical when each iteration is expensive, as is the case when dealing with modern data sets, such as genome-wide association studies with thousands of rows and up to hundreds of thousands of columns. We develop coupling techniques tailored to the setting of high-dimensional regression with shrinkage priors, which enable practical, non-asymptotic diagnostics of convergence without relying on traceplots or long-run asymptotics. By establishing geometric drift and minorization conditions for the algorithm under consideration, we prove that the proposed couplings have finite expected meeting time. Focusing on a class of shrinkage priors which includes the ‘Horseshoe’, we empirically demonstrate the scalability of the proposed couplings. A highlight of our findings is that less than 1000 iterations can be enough for a Gibbs sampler to reach stationarity in a regression on 100,000 covariates. The numerical results also illustrate the impact of the prior on the computational efficiency of the coupling, and suggest the use of priors where the local precisions are Half-t distributed with degree of freedom larger than one. We consider Markov chain Monte Carlo (MCMC) algorithms for Bayesian high-dimensional regression with "
]
"authors_fields" => array:2 [
"fr" => "Systèmes d'Information, Data Analytics et Opérations"
"en" => "Information Systems, Data Analytics and Operations"
]
"indexedAt" => "2025-03-16T10:21:40.000Z"
"docTitle" => "Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regression with shrinkage priors Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regressio "
"docSurtitle" => "Journal articles"
"authorNames" => "<a href="/cv/jacob-pierre">JACOB Pierre</a>, BISWAS Niloy, BHATTACHARYA Anirban, JOHNDROW James"
"docDescription" => "<span class="document-property-authors">JACOB Pierre, BISWAS Niloy, BHATTACHARYA Anirban, JOHNDROW James</span><br><span class="document-property-authors_fields">Information Systems, Data Analytics and Operations</span> | <span class="document-property-year">2022</span> <span class="document-property-authors">JACOB Pierre, BISWAS Niloy, BHATTACHARYA Anirban, JOHNDROW J "
"keywordList" => "<a href="#">Bayesian inference</a>, <a href="#">couplings</a>, <a href="#">Gibbs sampling</a>, <a href="#">Horseshoe prior</a>, <a href="#">parallel computation</a> <a href="#">Bayesian inference</a>, <a href="#">couplings</a>, <a href="#">Gibbs sampling</a>, <a hr "
"docPreview" => "<b>Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regression with shrinkage priors</b><br><span>2022-07 | Journal articles </span> <b>Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regres "
"docType" => "research"
"publicationLink" => "<a href="https://doi.org/10.1111/rssb.12495" target="_blank">Coupling-based convergence assessment of some Gibbs samplers for high-dimensional Bayesian regression with shrinkage priors</a> <a href="https://doi.org/10.1111/rssb.12495" target="_blank">Coupling-based convergence assessment o "
]
+lang: "en"
+"_type": "_doc"
+"_score": 8.989445
+"parent": null
}
