{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T21:52:37Z","timestamp":1791323557518,"version":"4.3.1"},"reference-count":46,"publisher":"Institute of Mathematical Statistics","issue":"3","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Ann. Statist."],"published-print":{"date-parts":[[2026,6,1]]},"abstract":"<jats:p>We derive novel and sharp high-dimensional Berry\u2013Esseen bounds for the sum of m-dependent random vectors over the class of hyperrectangles exhibiting only a poly-logarithmic dependence in the dimension. Our results hold under minimal assumptions, such as nondegenerate covariances and finite third moments, and exhibit an optimal sample complexity of order m(q\u22121)\/(q\u22122)\/n. Aside from logarithmic terms, the resulting rates match the optimal rates established in the univariate case. When specialized to the sums of independent nondegenerate random vectors, our results produce sharp and, in some cases, optimal rates under the weakest possible conditions. We develop a novel inductive relationship between anticoncentration inequalities and Berry\u2013Esseen bounds inspired by the classical Lindeberg swapping method and the concentration inequality approach for dependent data that may be of independent interest.<\/jats:p>","DOI":"10.1214\/25-aos2616","type":"journal-article","created":{"date-parts":[[2026,6,25]],"date-time":"2026-06-25T13:49:56Z","timestamp":1782395396000},"page":"1523-1539","source":"Crossref","is-referenced-by-count":0,"title":["Dual induction CLT for high-dimensional m-dependent data"],"prefix":"10.1214","volume":"54","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3906-2175","authenticated-orcid":false,"given":"Heejong","family":"Bong","sequence":"first","affiliation":[{"name":"Department of Statistics, Purdue University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4459-5352","authenticated-orcid":false,"given":"Arun Kumar","family":"Kuchibhotla","sequence":"additional","affiliation":[{"name":"Department of Statistics and Data Science, Carnegie Mellon University"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alessandro","family":"Rinaldo","sequence":"additional","affiliation":[{"name":"Department of Statistics and Data Science, University of Texas at Austin"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"108","reference":[{"key":"36","doi-asserted-by":"publisher","unstructured":"<small>Rai\u010d, M.<\/small> (2019). A multivariate Berry\u2013Esseen theorem with explicit constants. <i>Bernoulli<\/i> <b>25<\/b> 2824\u20132853.","DOI":"10.3150\/18-BEJ1072"},{"key":"45","doi-asserted-by":"publisher","unstructured":"<small>Zhang, X.<\/small> and <small>Cheng, G.<\/small> (2018). Gaussian approximation for high dimensional vector under physical dependence. <i>Bernoulli<\/i> <b>24<\/b> 2640\u20132675.","DOI":"10.3150\/17-BEJ939"},{"key":"42","doi-asserted-by":"publisher","unstructured":"<small>Zhang, D.<\/small> and <small>Wu, W. B.<\/small> (2017). Gaussian approximation for high dimensional time series. <i>Ann. Statist.<\/i> <b>45<\/b> 1895\u20131919.","DOI":"10.1214\/16-AOS1512"},{"key":"18","doi-asserted-by":"publisher","unstructured":"<small>DAS, D.<\/small> and <small>LAHIRI, S.<\/small> (2021). Central Limit Theorem in high dimensions: The optimal bound on dimension growth rate. <i>Trans. Amer. Math. Soc.<\/i> <b>374<\/b> 6991\u20137009.","DOI":"10.1090\/tran\/8459"},{"key":"19","doi-asserted-by":"publisher","unstructured":"<small>DENG, H.<\/small> and <small>ZHANG, C.-H.<\/small> (2020). Beyond Gaussian approximation: Bootstrap for maxima of sums of independent random vectors. <i>Ann. Statist.<\/i> <b>48<\/b> 3643\u20133671.","DOI":"10.1214\/20-AOS1946"},{"key":"21","doi-asserted-by":"publisher","unstructured":"<small>FANG, X.<\/small> and <small>KOIKE, Y.<\/small> (2021). High-dimensional central limit theorems by Stein\u2019s method. <i>Ann. Appl. Probab.<\/i> <b>31<\/b> 1660\u20131686.","DOI":"10.1214\/20-aap1629"},{"key":"31","doi-asserted-by":"publisher","unstructured":"<small>LOPES, M. E.<\/small>, <small>LIN, Z.<\/small> and <small>M\u00dcLLER, H.-G.<\/small> (2020). Bootstrapping max statistics in high dimensions: Near-parametric rates under weak variance decay and application to functional and multinomial data. <i>Ann. Statist.<\/i> <b>48<\/b> 1214\u20131229.","DOI":"10.1214\/19-AOS1844"},{"key":"46","doi-asserted-by":"publisher","unstructured":"<small>ZHILOVA, M.<\/small> (2020). Nonclassical Berry\u2013Esseen inequalities and accuracy of the bootstrap. <i>Ann. Statist.<\/i> <b>48<\/b> 1922\u20131939.","DOI":"10.1214\/18-AOS1802"},{"key":"7","doi-asserted-by":"publisher","unstructured":"<small>BERK, K. N.<\/small> (1973). A central limit theorem for <i>m<\/i>-dependent random variables with unbounded <i>m<\/i>. <i>Ann. Probab.<\/i> <b>1<\/b> 352\u2013354.","DOI":"10.1214\/aop\/1176996992"},{"key":"38","doi-asserted-by":"publisher","unstructured":"<small>ROMANO, J. P.<\/small> and <small>WOLF, M.<\/small> (2000). A more general central limit theorem for <i>m<\/i>-dependent random variables with unbounded <i>m<\/i>. <i>Statist. Probab. Lett.<\/i> <b>47<\/b> 115\u2013124.","DOI":"10.1016\/S0167-7152(99)00146-7"},{"key":"15","doi-asserted-by":"publisher","unstructured":"<small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small>, <small>KATO, K.<\/small> and <small>KOIKE, Y.<\/small> (2023). High-dimensional data bootstrap. <i>Annu. Rev. Stat. Appl.<\/i> <b>10<\/b> 427\u2013449.","DOI":"10.1146\/annurev-statistics-040120-022239"},{"key":"16","doi-asserted-by":"publisher","unstructured":"<small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small> and <small>KOIKE, Y.<\/small> (2023). Nearly optimal central limit theorem and bootstrap approximations in high dimensions. <i>Ann. Appl. Probab.<\/i> <b>33<\/b> 2374\u20132425.","DOI":"10.1214\/22-aap1870"},{"key":"24","doi-asserted-by":"publisher","unstructured":"<small>KOIKE, Y.<\/small> (2021). Notes on the dimension dependence in high-dimensional central limit theorems for hyperrectangles. <i>Jpn. J. Stat. Data Sci.<\/i> <b>4<\/b> 257\u2013297.","DOI":"10.1007\/s42081-020-00096-7"},{"key":"27","doi-asserted-by":"publisher","unstructured":"<small>KUCHIBHOTLA, A. K.<\/small>, <small>MUKHERJEE, S.<\/small> and <small>BANERJEE, D.<\/small> (2021). High-dimensional CLT: Improvements, non-uniform extensions and large deviations. <i>Bernoulli<\/i> <b>27<\/b> 192\u2013217.","DOI":"10.3150\/20-BEJ1233"},{"key":"1","doi-asserted-by":"crossref","unstructured":"<small>Andrews, D. W. K.<\/small> (1991). Heteroskedasticity and autocorrelation consistent covariance matrix estimation. <i>Econometrica<\/i> <b>59<\/b> 817\u2013858.","DOI":"10.2307\/2938229"},{"key":"10","doi-asserted-by":"publisher","unstructured":"<small>Chen, L. H. Y.<\/small> and <small>Shao, Q.-M.<\/small> (2004). Normal approximation under local dependence. <i>Ann. Probab.<\/i> <b>32<\/b> 1985\u20132028.","DOI":"10.1214\/009117904000000450"},{"key":"11","doi-asserted-by":"publisher","unstructured":"<small>Chernozhukov, V.<\/small>, <small>Chetverikov, D.<\/small> and <small>Kato, K.<\/small> (2013). Gaussian approximations and multiplier bootstrap for maxima of sums of high-dimensional random vectors. <i>Ann. Statist.<\/i> <b>41<\/b> 2786\u20132819.","DOI":"10.1214\/13-AOS1161"},{"key":"14","doi-asserted-by":"publisher","unstructured":"<small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small>, <small>KATO, K.<\/small> and <small>KOIKE, Y.<\/small> (2022). Improved central limit theorem and bootstrap approximations in high dimensions. <i>Ann. Statist.<\/i> <b>50<\/b> 2562\u20132586.","DOI":"10.1214\/22-aos2193"},{"key":"25","doi-asserted-by":"publisher","unstructured":"<small>KOIKE, Y.<\/small> (2023). High-dimensional central limit theorems for homogeneous sums. <i>J. Theoret. Probab.<\/i> <b>36<\/b> 1\u201345.","DOI":"10.1007\/s10959-022-01156-2"},{"key":"2","doi-asserted-by":"crossref","unstructured":"<small>BELLONI, A.<\/small>, <small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small>, <small>HANSEN, C. B.<\/small> and <small>KATO, K.<\/small> (2018). High-dimensional econometrics and regularized GMM cemmap working paper No. CWP35\/18, Centre for Microdata Methods and Practice (cemmap), London. <a href=\"https:\/\/doi.org\/10.1920\/wp.cem.2018.3518\">https:\/\/doi.org\/10.1920\/wp.cem.2018.3518<\/a>.","DOI":"10.1920\/wp.cem.2018.3518"},{"key":"3","doi-asserted-by":"publisher","unstructured":"<small>BENTKUS, V.<\/small> (2003). Dependence of the Berry\u2013Esseen bound on the dimension. <i>J. Statist. Plann. Inference<\/i> <b>113<\/b> 385\u2013402.","DOI":"10.1016\/S0378-3758(02)00094-0"},{"key":"4","doi-asserted-by":"publisher","unstructured":"<small>BENTKUS, V.<\/small> (2005). A Lyapunov-type bound in <math xmlns=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" overflow=\"scroll\" id=\"eq_2\"><msup><mrow><mi mathvariant=\"double-struck\">R<\/mi><\/mrow><mrow><mi mathvariant=\"italic\">d<\/mi><\/mrow><\/msup><\/math>. <i>Theory Probab. Appl.<\/i> <b>49<\/b> 311\u2013323.","DOI":"10.1137\/S0040585X97981123"},{"key":"5","doi-asserted-by":"publisher","unstructured":"<small>BENTKUS, V.<\/small>, <small>G\u00d6TZE, F.<\/small>, <small>PAULAUSKAS, V.<\/small> and <small>RA\u010cKAUSKAS, A.<\/small> (2000). The accuracy of Gaussian approximation in Banach spaces. In <i>Limit Theorems of Probability Theory<\/i> (Y. V. Prokhorov and V. Statulevi\u010dius, eds.) 25\u2013111. Springer, Berlin.","DOI":"10.1007\/978-3-662-04172-7_2"},{"key":"6","doi-asserted-by":"publisher","unstructured":"<small>BENTKUS, V. Y.<\/small> (1987). Large deviations in Banach spaces. <i>Theory Probab. Appl.<\/i> <b>31<\/b> 627\u2013632.","DOI":"10.1137\/1131085"},{"key":"8","doi-asserted-by":"crossref","unstructured":"<small>BONG, H.<\/small>, <small>KUCHIBHOTLA, A. K.<\/small> and <small>RINALDO, A.<\/small> (2026). Supplement to \u201cDual induction CLT for high-dimensional <i>m<\/i>-dependent data.\u201d <a href=\"https:\/\/doi.org\/10.1214\/25-AOS2616SUPP\">https:\/\/doi.org\/10.1214\/25-AOS2616SUPP<\/a>","DOI":"10.1214\/25-AOS2616SUPP"},{"key":"13","doi-asserted-by":"publisher","unstructured":"<small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small> and <small>KATO, K.<\/small> (2018). Inference on causal and structural parameters using many moment inequalities. <i>Rev. Econ. Stud.<\/i> <b>86<\/b> 1867\u20131900.","DOI":"10.1093\/restud\/rdy065"},{"key":"17","doi-asserted-by":"publisher","unstructured":"<small>CHIANG, H. D.<\/small>, <small>KATO, K.<\/small> and <small>SASAKI, Y.<\/small> (2023). Inference for high-dimensional exchangeable arrays. <i>J. Amer. Statist. Assoc.<\/i> <b>118<\/b> 1595\u20131605.","DOI":"10.1080\/01621459.2021.2000868"},{"key":"20","unstructured":"<small>FANG, X.<\/small> and <small>KOIKE, Y.<\/small> (2020). Large-dimensional central limit theorem with fourth-moment error bounds on convex sets and balls. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/2009.00339\">2009.00339<\/a>."},{"key":"22","unstructured":"<small>FANG, X.<\/small>, <small>KOIKE, Y.<\/small>, <small>LIU, S.-H.<\/small> and <small>YI-KUN, Z.<\/small> (2023). High-dimensional central limit theorems by Stein\u2019s method in the degenerate case. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/2305.17365\">2305.17365<\/a>."},{"key":"23","unstructured":"<small>KOCK, A. B.<\/small> and <small>PREINERSTORFER, D.<\/small> (2025). High-dimensional Gaussian approximations for robust means. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/2504.08435\">2504.08435<\/a>."},{"key":"26","doi-asserted-by":"publisher","unstructured":"<small>KOJEVNIKOV, D.<\/small> and <small>SONG, K.<\/small> (2022). A Berry\u2013Esseen bound for vector-valued martingales. <i>Statist. Probab. Lett.<\/i> <b>186<\/b> 109448.","DOI":"10.1016\/j.spl.2022.109448"},{"key":"28","unstructured":"<small>KUCHIBHOTLA, A. K.<\/small> and <small>RINALDO, A.<\/small> (2020). High-dimensional CLT for sums of non-degenerate random Vectors. <math xmlns=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" overflow=\"scroll\" id=\"eq_3\"><msup><mrow><mi mathvariant=\"italic\">n<\/mi><\/mrow><mrow><mo>\u2212<\/mo><mo fence=\"true\" stretchy=\"false\">(<\/mo><mn>1<\/mn><mo stretchy=\"false\">\/<\/mo><mn>2<\/mn><mo fence=\"true\" stretchy=\"false\">)<\/mo><\/mrow><\/msup><\/math>-rate. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/2009.13673\">2009.13673<\/a>."},{"key":"29","doi-asserted-by":"publisher","unstructured":"<small>KURISU, D.<\/small>, <small>KATO, K.<\/small> and <small>SHAO, X.<\/small> (2024). Gaussian approximation and spatially dependent wild bootstrap for high-dimensional spatial data. <i>J. Amer. Statist. Assoc.<\/i> <b>119<\/b> 1820\u20131832.","DOI":"10.1080\/01621459.2023.2218578"},{"key":"30","doi-asserted-by":"publisher","unstructured":"<small>LOPES, M. E.<\/small> (2022). Central limit theorem and bootstrap approximation in high dimensions: Near <math xmlns=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" overflow=\"scroll\" id=\"eq_4\"><mn>1<\/mn><mo stretchy=\"false\">\/<\/mo><msqrt><mrow><mi mathvariant=\"italic\">n<\/mi><\/mrow><\/msqrt><\/math> rates via implicit smoothing. <i>Ann. Statist.<\/i> <b>50<\/b> 2492\u20132513.","DOI":"10.1214\/22-AOS2184"},{"key":"32","doi-asserted-by":"publisher","unstructured":"<small>NAZAROV, F.<\/small> (2003). On the maximal perimeter of a convex set in <math xmlns=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" overflow=\"scroll\" id=\"eq_5\"><msup><mrow><mi mathvariant=\"double-struck\">R<\/mi><\/mrow><mrow><mi mathvariant=\"italic\">n<\/mi><\/mrow><\/msup><\/math> with respect to a Gaussian measure. In <i>Geometric Aspects of Functional Analysis<\/i>: <i>Israel Seminar<\/i> 2001<i>\u2013<\/i>2002 (V. D. Milman and G. Schechtman, eds.) 169\u2013187. Springer, Berlin.","DOI":"10.1007\/978-3-540-36428-3_15"},{"key":"33","doi-asserted-by":"publisher","unstructured":"<small>NEWEY, W. K.<\/small> and <small>WEST, K. D.<\/small> (1987). A simple, positive semi-definite, heteroskedasticity and autocorrelation consistent covariance matrix. <i>Econometrica<\/i> <b>55<\/b> 703\u2013708.","DOI":"10.2307\/1913610"},{"key":"34","doi-asserted-by":"publisher","unstructured":"<small>NORVAI\u0160A, R.<\/small> and <small>PAULAUSKAS, V.<\/small> (1991). Rate of convergence in the central limit theorem for empirical processes. <i>J. Theoret. Probab.<\/i> <b>4<\/b> 511\u2013534.","DOI":"10.1007\/BF01210322"},{"key":"35","doi-asserted-by":"publisher","unstructured":"<small>PAULAUSKAS, V.<\/small> and <small>RA\u010cKAUSKAS, A. A.<\/small> (2012). <i>Approximation Theory in the Central Limit Theorem<\/i>: <i>Exact Results in Banach Spaces<\/i> 32. Springer, Berlin.","DOI":"10.1007\/978-94-011-7798-6"},{"key":"37","doi-asserted-by":"publisher","unstructured":"<small>RINOTT, Y.<\/small> and <small>ROTAR, V.<\/small> (1999). Some bounds on the rate of convergence in the CLT for martingales. I. <i>Theory Probab. Appl.<\/i> <b>43<\/b> 604\u2013619.","DOI":"10.1137\/S0040585X97977148"},{"key":"39","doi-asserted-by":"crossref","unstructured":"<small>SAZONOV, V. V.<\/small> (1981). <i>Normal Approximation\u2014Some Recent Advances<\/i>. <i>Lecture Notes in Math.<\/i> <b>879<\/b>. Springer, Berlin.","DOI":"10.1007\/BFb0096862"},{"key":"40","doi-asserted-by":"publisher","unstructured":"<small>SENATOV, V. V.<\/small> (1998). <i>Normal Approximation<\/i>: <i>New Results, Methods and Problems<\/i>. de Gruyter, Berlin.","DOI":"10.1515\/9783110933666"},{"key":"41","doi-asserted-by":"publisher","unstructured":"<small>SHERGIN, V. V.<\/small> (1980). On the convergence rate in the central limit theorem for m-dependent random variables. <i>Theory Probab. Appl.<\/i> <b>24<\/b> 782\u2013796.","DOI":"10.1137\/1124090"},{"key":"43","unstructured":"<small>ZHANG, H.<\/small> and <small>CHEN, S. X.<\/small> (2020). Concentration inequalities for statistical inference. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/2011.02258\">2011.02258<\/a>."},{"key":"44","unstructured":"<small>ZHANG, X.<\/small> and <small>CHENG, G.<\/small> (2014). Bootstrapping high dimensional time series. arXiv preprint. Available at arXiv:<a href=\"https:\/\/arxiv.org\/abs\/1406.1037\">1406.1037<\/a>."},{"key":"9","doi-asserted-by":"publisher","unstructured":"<small>CHANG, J.<\/small>, <small>CHEN, X.<\/small> and <small>WU, M.<\/small> (2024). Central limit theorems for high dimensional dependent data. <i>Bernoulli<\/i> <b>30<\/b> 712\u2013742.","DOI":"10.3150\/23-bej1614"},{"key":"12","doi-asserted-by":"publisher","unstructured":"<small>CHERNOZHUKOV, V.<\/small>, <small>CHETVERIKOV, D.<\/small> and <small>KATO, K.<\/small> (2017). Central limit theorems and bootstrap in high dimensions. <i>Ann. Probab.<\/i> <b>45<\/b> 2309\u20132352.","DOI":"10.1214\/16-AOP1113"}],"container-title":["The Annals of Statistics"],"original-title":[],"deposited":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T21:19:36Z","timestamp":1791321576000},"score":1,"resource":{"primary":{"URL":"https:\/\/projecteuclid.org\/journals\/annals-of-statistics\/volume-54\/issue-3\/Dual-induction-CLT-for-high-dimensional-m-dependent-data\/10.1214\/25-AOS2616.full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,1]]},"references-count":46,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,6,1]]}},"URL":"https:\/\/doi.org\/10.1214\/25-aos2616","relation":{},"ISSN":["0090-5364","2168-8966"],"issn-type":[{"value":"0090-5364","type":"print"},{"value":"2168-8966","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,1]]}}}