iiinova — where your digital self keeps evolving

Statistics & Probability Letters

@spletters.bsky.social

59 followers 40 following 41 posts

Statistics & Probability Letters publishes concise articles covering research findings in statistics and probability. Bluesky social of Statistics & Probability Letters

Posts Replies Media

Statistics & Probability Letters @spletters.bsky.social · 14/09/2026

❗Retrieval-augmented generation theory❗ A new statistical proxy framework explains when retrieval can be trusted & analyzes hallucinations via retrieval discordance www.sciencedirect.com/science/arti... By G Biau and C Boye @iufrance.bsky.social @sorbonne-universite.fr @univparissaclay.bsky.social

sciencedirect.com

Statistics & Probability Letters @spletters.bsky.social · 19/08/2026

❗New paper alert❗ 👉K. Kawamoto, Y. Goto, and K. Tsukuda (Kyushu University) study spectral clustering based on the sign of the first principal component score under a two-component Gaussian mixture model with general covariance matrices. Read the letter⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

Statistics & Probability Letters @spletters.bsky.social · 12/08/2026

❗New paper alert ❄️In our December issue: 👉Svyatoslav Novikov (@unil.bsky.social) studies the asymptotic probability that a Gaussian field exceeds a high level when its variance structure contains a product-type term. Read the letter here ⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

Extremes of Gaussian fields with a product term in the variance

We study the high excursion probability of a centered Gaussian field on a square. Writing σ and r for its standard deviation and correlation function,…

Statistics & Probability Letters @spletters.bsky.social · 06/08/2026

Shintaro Hashimoto (Hiroshima University) derives objective priors that asymptotically match the mean of the Bayesian predictive distribution with that of the frequentist plug-in predictive distribution. Find all the details in our November issue! www.sciencedirect.com/science/arti...

sciencedirect.com

Statistics & Probability Letters @spletters.bsky.social · 23/07/2026

❗New robust tool for survival analysis w complex covariate effects & heterogeneous variability: ➡️Jinho Park (Inha University) proposes a method for quantile regression in a reproducing kernel Hilbert space for censored data. 🔗https://www.sciencedirect.com/science/article/abs/pii/S0167715226001884

Statistics & Probability Letters @spletters.bsky.social · 13/07/2026

M. Allouche (Kaiko) & E. Gobet (Sorbonne University) establish a multivariate CLT for weighted empirical quantiles and extend the classical Wilks method to derive an asymptotically distribution-free confidence interval for weighted empirical quantiles. ⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 01/06/2026

Jianhang Ai (Czech Technical University) & Christos Pelekis (Aristotle University of Thessaloniki) deduce a lower bound on the probability that a hypergeometric random variable exceeds its expected value, in two different parameter-regimes. In our Sept issue⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

On lower bounds for hypergeometric tails

Let n,k be positive integers such that n≥k, and let H be a hypergeometric random variable counting the number of black marbles in a sample without rep…

Statistics & Probability Letters @spletters.bsky.social · 28/05/2026

Pandey & Selvamuthu (Indian Institute of Technology Delhi) introduce and study the marked dynamic contagion process, where the intensity process is governed by the Cox–Ingersoll–Ross, extending the Hawkes and the Cox process with Poisson shot-noise intensity. 🔗 www.sciencedirect.com/science/arti...

sciencedirect.com

Limit theorems for marked dynamic contagion processes with mean-reverting Cox–Ingersoll–Ross intensity

In this paper, we introduce and study the marked dynamic contagion process, where the intensity process is governed by the Cox–Ingersoll–Ross (CIR) mo…

Statistics & Probability Letters @spletters.bsky.social · 08/04/2026

@lambertdem.bsky.social and Ioannis Papastathopoulos (@edinburgh-uni.bsky.social) show the convergence of directionally-renormalised univariate random point measures to a limiting directional Poisson point process for small and large extremes. Read the letter⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 02/04/2026

@xbardina.bsky.social & Salim Boukfal (@uab.cat) generalize the known result of Kac and Stroock, showing that processes derived from renewal processes, whose inter-arrival times satisfy weak integrability conditions, weakly converge to Brownian motion. ⤵️⤵️⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

Kac–Stroock type approximations for the Brownian motion from renewal processes

In the present paper we show that the processes Xn={Xn(t):t∈[0,1]}, n∈N, defined by Xn(t)=nC∫0t(−1)L(nu)du, where L={L(t):t≥0} is a renewal process wh…

Statistics & Probability Letters @spletters.bsky.social · 26/03/2026

Robert Gaunt (University of Manchester) obtains a closed-form formula for the characteristic function of the asymmetric Student's t-distribution, also establishing a closed-form formula for a natural integral involving the sine function. Check this out⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 18/03/2026

JB Hill (UNC-ChapelHill) extends the property of symmetrization to dependent r.v.'s in an arbitrary setting, without requiring dependence or heterogeneity conditions. Usual bounds yielded under symmetrization are now functions of high dimensional Gaussian approximations. tinyurl.com/y8569fdy

sciencedirect.com

Symmetrization for high dimensional dependent random variables

We establish a generic symmetrization property for dependent random variables {xt}t=1n on Rp, where p ≫ n is allowed. We link Eψ(max1≤i≤p|1/n∑t=1n(xi,…

Statistics & Probability Letters @spletters.bsky.social · 12/03/2026

A. Silbernagel & C. H. Weiß (Helmut Schmidt University) propose a refined definition of the INAR random field model, provide a derivation of its autocorrelation structure, and expand the model to higher-order autoregressions and study its stochastic properties⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 04/03/2026

Lea Maislinger & Wolfgang Trutschnig (University of Salzburg) show that even supposedly handy and regular classes like the family of multivariate Archimedean copulas may contain pathological (but interesting) elements. Find out more in our July issue⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 05/02/2026

Djordjilović (Ca' Foscari), Sofer (Beth Israel) & Dreyfuss (Joslin Diabetes Center) study the directional replicability of an effect across independent studies providing conditions under which one can combine 2 one-sided replicability tests w/o correcting for multiplicity. tinyurl.com/34wpj6u5

Statistics & Probability Letters @spletters.bsky.social · 28/01/2026

@piergiovannib.bsky.social & @matteoborrotti.bsky.social (UNIMIB) study uncertainty quantification via martingale posteriors in generative classification models. To this end, the authors consider a conditionally identically distributed sequence of observations. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 23/01/2026

In our April issue, Lorenzo Camponovo (@usi.ch & SUPSI) studies the relationship between stochastic and deterministic polynomial trends over the long-run, as the sample size increases to infinity, and at a local level. Read the letter here⤵️ www.sciencedirect.com/science/arti...

This link will take you to a page that’s not on LinkedIn

Statistics & Probability Letters @spletters.bsky.social · 14/01/2026

✨ Spotlight Papers of 2025, II Semester✨ In case you missed them⤵️ 🔗 www.sciencedirect.com/science/arti... by S. Zwanzig & R. Ahmad 🔗 www.sciencedirect.com/science/arti... by M. Hua & G. Goh

Statistics & Probability Letters @spletters.bsky.social · 09/01/2026

HNY! Our Co-Editor-in-Chief @afarcome.bsky.social elected the ✨ Spotlight paper of the First semester of 2025✨ "Universal distribution of the empirical coverage in split conformal prediction" by Paulo C. Marques F. (Insper Instituto de Ensino e Pesquisa) www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 15/12/2025

Swarnadeep Datta (Indian Statistical Institute) and Monitirtha Dey (@unibremen.bsky.social) propose a simple single-step multiple testing procedure that asymptotically controls the FWER at the desired level exactly under the equicorrelated multivariate Gaussian setup. doi.org/10.1016/j.sp...

Statistics & Probability Letters @spletters.bsky.social · 10/12/2025

@luisantonioarteaga.bsky.social & JM Rodriguez-Poo (Univ de Cantabria) propose a new test to detect structural changes in economic models with endogenous variables and time dependence using nonparametric techniques that avoid the curse of dimensionality. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 01/12/2025

C Scricciolo (Univ of Verona) establishes minimax-optimal convergence rates in the 1-Wasserstein metric for a kernel-type estimator of the CDF + proposes a data-driven adaptive estimation procedure to select the optimal bandwidth for α-Hölder mixing densities www.sciencedirect.com/science/arti...

sciencedirect.com

Adaptive minimax-optimal Wasserstein deconvolution with unknown error distributions

We study the problem of deconvolving an unknown distribution function when the error distribution is ordinary smooth and unknown. Using data from an a…

Statistics & Probability Letters @spletters.bsky.social · 18/11/2025

Manabu Asai (Soka University) utilizes the Bartlett decomposition of a singular Wishart variable to derive the log-expectation and proposes a reparameterization for MLE to separately estimate the scale matrix and the degrees-of-freedom parameter. ⤵️⤵️⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 05/11/2025

Marco Geraci (Sapienza University of Rome) explores an easy-to-implement maximum likelihood estimation strategy based on Gaussian quadrature for the scale-mixture representation of the generalized Laplace distribution and its projection onto the circle. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 15/10/2025

Continuous monitoring is gaining popularity for its ability to reduce sample sizes and reach earlier conclusions. Long-Hao Xu and Tim Friede (Universitätsmedizin Göttingen) investigate the theoretical properties of blinded continuous monitoring. www.sciencedirect.com/science/arti...

sciencedirect.com

A note on blinded continuous monitoring for continuous outcomes

Continuous monitoring is becoming more popular due to its significant benefits, including reducing sample sizes and reaching earlier conclusions. In g…

Statistics & Probability Letters @spletters.bsky.social · 08/10/2025

🚨New letter alert 🚨 Hang Zou (Guangdong University of Foreign Studies) and Yunlu Jiang (Jinan University) propose a communication-efficient distributed robust variable selection method using discounted exponential regression for massive data. www.sciencedirect.com/science/arti...

sciencedirect.com

Communication-efficient distributed robust variable selection for heterogeneous massive data

We propose a communication-efficient distributed robust variable selection method using discounted exponential regression for massive data. Theoretica…

Statistics & Probability Letters @spletters.bsky.social · 26/09/2025

Yves Tillé (Uni of Neuchâtel) introduces a 2-phase procedure for balanced random sample selection. ✈️Flight phase: transform the vector of inclusion probabilities into one with components close to 0 or 1, 🛬 Landing phase: Complete the selection. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 15/09/2025

In our November issue: S. Zwanzig & R. Ahmad (Uppsala University) show that in a linear model setting, the minimization problem in variational inference pertains to approximation error under a suitable prior, and discuss the choice of prior. www.sciencedirect.com/science/arti...

sciencedirect.com

Statistics & Probability Letters @spletters.bsky.social · 09/09/2025

Swagatam Das (Indian Statistical Institute) proposes a free probabilistic framework to analyze Transformer-based LLMs, emphasizing the role of non-commutative random variables in modeling word order and contextual structure. In our November issue⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 29/08/2025

Eunji Lim (Adelphi University) proposes a penalized estimator to address boundary overfitting in isotonic regression. The estimator achieves strong uniform consistency and ensures consistency at domain boundaries. In our November issue⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 26/08/2025

There are a few options for clustering mixed-type data. Eman Alamer (McMaster University), Michael Gallaugher (Baylor University), and Paul McNicholas (McMaster University) propose a model-based clustering approach to help address this gap. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 11/08/2025

@alecasa.bsky.social and Davide Ferrari (@unibz.bsky.social) tackle the problem of selecting the number of components in Gaussian mixture models, where traditional single-model criteria like BIC often overlook model selection uncertainty. Learn more⤵️ www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 31/07/2025

Seonghyun Jeong (Yonsei University) derives a desirable test function that establishes a framework for testing-based posterior contraction theory in Gaussian models with unknown variance. Read the letter in our November issue⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

L2-norm posterior contraction in Gaussian models with unknown variance

The testing-based approach is a fundamental tool for establishing posterior contraction rates. Although the Hellinger metric is attractive owing to th…

Statistics & Probability Letters @spletters.bsky.social · 24/07/2025

Lu Mao (Univ. of Wisconsin-Madison) approximates win-loss statistics using only the marginal survival curves typically reported in clinical trials, enabling comparative interpretation of two-tiered outcomes even when joint data are unavailable #meta-analysis 🔗 www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 16/07/2025

Hervé Cardot (Université Bourgogne Europe) and Antonio Musolesi (University of Ferrara and SEEDS) propose a zero-inflated model for the temporal variation of the variable of interest in panel data analysis. Click below to learn more⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

Partially time-invariant panel data regression

In panel data analysis, temporal variation in the variable of interest is commonly exploited to eliminate individual-specific effects. However, even w…

Statistics & Probability Letters @spletters.bsky.social · 10/07/2025

A Ghoshal and JT Chen (Bowling Green State U) propose a sequentially rejective procedure for simultaneous inference on odds ratios, which is uniformly more powerful than Holm’s procedure when testing multiple odds ratios. Read the letter in our November issue🔗 www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 02/07/2025

🍂 In our September issue Yves Tillé (University of Neuchâtel) proposes a simple bootstrap method for large-entropy, unequal-probability sampling designs in finite populations. An efficient implementation in R is also provided. www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 28/05/2025

🍂In our September issue🍂 Masanari Kimura (University of Melbourne) develops a higher-order asymptotic analysis for the semi-hard triplet loss using the Edgeworth expansion. 🔗Click here to read the letter www.sciencedirect.com/science/arti...

Statistics & Probability Letters @spletters.bsky.social · 21/05/2025

🍃In our September issue🍃 The Marchenko–Pastur law describes the limit law of eigenvalues of large rectangular matrices. Luc Devroye (@lawmcgill.bsky.social) gives two efficient algorithms for simulating random variables from this distribution. Read the paper⤵️ www.sciencedirect.com/science/arti...

sciencedirect.com

Exact simulation of the Marchenko–Pastur distribution

The Marchenko–Pastur law (Marchenko and Pastur, 1967) describes the limit law of eigenvalues of large rectangular matrices. We give two efficient algo…

Statistics & Probability Letters @spletters.bsky.social · 30/04/2025

🚨 In our July issue 🚨 L Li (Nankai Uni), Y Yu and W Liang (Dongbei Uni of Finance and Economics), and F Zou (Zhongnan Uni of Economics and Law) study multi-attribute Gaussian copula graphical models where each node captures multidimensional variables 🔍 www.sciencedirect.com/science/arti...

sciencedirect.com

A novel approach for estimating multi-attribute Gaussian copula graphical models

This paper considers estimating multi-attribute Gaussian copula graphical models where each node represents multivariate variables with rich meanings.…

Statistics & Probability Letters @spletters.bsky.social · 25/04/2025

SPL has landed on Bluesky 🦋 We are thrilled to join the community and share the latest "Statistics & Probabiliy Letters" Stay tuned for new research insights!