{"id":1003,"date":"2025-12-23T16:09:41","date_gmt":"2025-12-23T16:09:41","guid":{"rendered":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/?p=1003"},"modified":"2025-12-24T12:47:16","modified_gmt":"2025-12-24T12:47:16","slug":"decoupling-for-markov-chains","status":"publish","type":"post","link":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/","title":{"rendered":"Decoupling for Markov Chains"},"content":{"rendered":"<p><strong>How can we rigorously quantify Monte Carlo error and assess convergence in modern MCMC methods such as the No-U-Turn Sampler?<\/strong><\/p>\n<p>This question motivates <a href=\"https:\/\/arxiv.org\/abs\/2512.19351\">new joint work<\/a> with <a href=\"https:\/\/www.columbia.edu\/~vhd1\/\">Victor de la Pe\u00f1a<\/a> (Columbia), which introduces a decoupling-based perspective for Markov chains with direct applications to MCMC.<\/p>\n<p><strong>Core idea.<\/strong><\/p>\n<p>Essentially all MCMC algorithms &#8212; from Gibbs samplers to Hamiltonian Monte Carlo and NUTS &#8212; can be written as deterministic update maps driven by i.i.d. auxiliary random variables.\u00a0 While this representation is well known, we show that it can be exploited in a systematic and largely unexplored way.<\/p>\n<p>The paper introduces a <strong>tangent-decoupling framework for Markov chains<\/strong>.\u00a0Alongside the original chain, we construct a companion process by re-running the same update map along the realized trajectory, but with fresh auxiliary randomness injected at each step. The resulting tangent-decoupled sequence<\/p>\n<ul>\n<li class=\"p1\">is straightforward to simulate in tandem with the original chain,<\/li>\n<li class=\"p1\">is conditionally independent given the realized backbone trajectory, and<\/li>\n<li class=\"p1\">is generally non-Markovian, yet remains ergodically correct.<\/li>\n<\/ul>\n<p class=\"p1\"><b>Main results.<\/b><b><\/b><\/p>\n<p class=\"p1\">Two theoretical consequences of this construction are particularly striking.<\/p>\n<p class=\"p1\">First, we establish <i>almost sure consistency of empirical averages<\/i> computed from the tangent-decoupled sequence, despite its lack of Markovian structure. Under mild conditions, these averages converge to the correct target expectation, yielding a structurally simple and broadly applicable estimator.<\/p>\n<p class=\"p1\">Second, we prove a <i>sharp, nonasymptotic variance inequality<\/i>: for any square-integrable observable and any finite sample size, the variance of the standard MCMC estimator is bounded above by twice the variance of the tangent-decoupled estimator. This bound is assumption-free: it does not rely on reversibility, spectral gaps, or mixing-time arguments; and leads directly to principled uncertainty quantification for MCMC output.<\/p>\n<p><strong>Why the connection is interesting mathematically<\/strong><\/p>\n<p>Decoupling theory is a classical tool in probability, with deep connections to martingale theory, empirical processes, and the study of dependent random variables. At its core, decoupling provides a principled way to compare dependent processes with suitably constructed independent counterparts.<\/p>\n<p>What is novel here is that this abstract theory is brought into direct contact with the concrete structure of Markov chains. By identifying a natural tangent-decoupled companion process, the work shows how general decoupling principles can be realized in an explicit, algorithmically meaningful setting.<\/p>\n<p class=\"p1\">Beyond the specific results, this perspective suggests a new way to think about convergence diagnostics and error assessment in MCMC algorithms, yielding tools for error assessment that remain valid even when classical Markov-chain assumptions are unavailable or hard to verify.<\/p>\n<blockquote><p>I am very grateful to Victor for this collaboration. The project was sparked by a question he raised after an applied probability seminar I gave at Columbia, a reminder of the well-known adage that <i>good questions are often the engine of good mathematics<\/i>, and highlighting the role of seminars not only as venues for dissemination, but as catalysts for new ideas and collaborations.<\/p><\/blockquote>\n<p>The full paper is available here:\u00a0<a href=\"https:\/\/arxiv.org\/abs\/2512.19351\">https:\/\/arxiv.org\/abs\/2512.19351<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>How can we rigorously quantify Monte Carlo error and assess convergence in modern MCMC methods such as the No-U-Turn Sampler? This question motivates new joint work with Victor de la &hellip; <a href=\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\" class=\"\">Read More<\/a><\/p>\n","protected":false},"author":2614,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[8],"tags":[],"class_list":["post-1003","post","type-post","status-publish","format-standard","hentry","category-preprints"],"acf":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v23.5 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Decoupling for Markov Chains - Nawaf Bou-Rabee<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Decoupling for Markov Chains - Nawaf Bou-Rabee\" \/>\n<meta property=\"og:description\" content=\"How can we rigorously quantify Monte Carlo error and assess convergence in modern MCMC methods such as the No-U-Turn Sampler? This question motivates new joint work with Victor de la &hellip; Read More\" \/>\n<meta property=\"og:url\" content=\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\" \/>\n<meta property=\"og:site_name\" content=\"Nawaf Bou-Rabee\" \/>\n<meta property=\"article:published_time\" content=\"2025-12-23T16:09:41+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-12-24T12:47:16+00:00\" \/>\n<meta name=\"author\" content=\"Nawaf Bou-Rabee\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Nawaf Bou-Rabee\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"2 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\",\"url\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\",\"name\":\"Decoupling for Markov Chains - Nawaf Bou-Rabee\",\"isPartOf\":{\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#website\"},\"datePublished\":\"2025-12-23T16:09:41+00:00\",\"dateModified\":\"2025-12-24T12:47:16+00:00\",\"author\":{\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/f9a7363ddc14a33eef77b520728dbde0\"},\"breadcrumb\":{\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Decoupling for Markov Chains\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#website\",\"url\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/\",\"name\":\"Nawaf Bou-Rabee\",\"description\":\"\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/f9a7363ddc14a33eef77b520728dbde0\",\"name\":\"Nawaf Bou-Rabee\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/f3f33e8b5e6e2430997aea3dfec1a454?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/f3f33e8b5e6e2430997aea3dfec1a454?s=96&d=mm&r=g\",\"caption\":\"Nawaf Bou-Rabee\"},\"description\":\"Math Professor at Rutgers\",\"sameAs\":[\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/\"],\"url\":\"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/author\/nb361\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Decoupling for Markov Chains - Nawaf Bou-Rabee","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/","og_locale":"en_US","og_type":"article","og_title":"Decoupling for Markov Chains - Nawaf Bou-Rabee","og_description":"How can we rigorously quantify Monte Carlo error and assess convergence in modern MCMC methods such as the No-U-Turn Sampler? This question motivates new joint work with Victor de la &hellip; Read More","og_url":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/","og_site_name":"Nawaf Bou-Rabee","article_published_time":"2025-12-23T16:09:41+00:00","article_modified_time":"2025-12-24T12:47:16+00:00","author":"Nawaf Bou-Rabee","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Nawaf Bou-Rabee","Est. reading time":"2 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/","url":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/","name":"Decoupling for Markov Chains - Nawaf Bou-Rabee","isPartOf":{"@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#website"},"datePublished":"2025-12-23T16:09:41+00:00","dateModified":"2025-12-24T12:47:16+00:00","author":{"@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/f9a7363ddc14a33eef77b520728dbde0"},"breadcrumb":{"@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/decoupling-for-markov-chains\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/"},{"@type":"ListItem","position":2,"name":"Decoupling for Markov Chains"}]},{"@type":"WebSite","@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#website","url":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/","name":"Nawaf Bou-Rabee","description":"","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/f9a7363ddc14a33eef77b520728dbde0","name":"Nawaf Bou-Rabee","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/f3f33e8b5e6e2430997aea3dfec1a454?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/f3f33e8b5e6e2430997aea3dfec1a454?s=96&d=mm&r=g","caption":"Nawaf Bou-Rabee"},"description":"Math Professor at Rutgers","sameAs":["https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/"],"url":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/author\/nb361\/"}]}},"_links":{"self":[{"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/posts\/1003"}],"collection":[{"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/users\/2614"}],"replies":[{"embeddable":true,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/comments?post=1003"}],"version-history":[{"count":5,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/posts\/1003\/revisions"}],"predecessor-version":[{"id":1008,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/posts\/1003\/revisions\/1008"}],"wp:attachment":[{"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/media?parent=1003"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/categories?post=1003"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/sites.rutgers.edu\/nawaf-bou-rabee\/wp-json\/wp\/v2\/tags?post=1003"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}