SearcharxivSearch

arXiv subjects

Nicolas Escobar-Velasquez

Publications and source records attributed to Nicolas Escobar-Velasquez.

4 recordsLinked to original sources

Pseudo-Differential Operators and Generalized Random Fields over Tori

Matérn covariance functions are ubiquitous in spatial statistics, valued for their interpretable parameters and well-understood sample path properties in Euclidean settings. This paper extends the theory of Matérn fields over tori through the pseudo-differential calculus. We first establish that a Matérn process on the $d$-dimensional torus has sample paths in $C^{ν^-}_{\mathrm{loc}}$ for every $ν>0$ and no more, exactly as in the Euclidean case. Our main results concern symbols whose order varies with position. Given an order function $m \in C^\infty(\mathbb{T}^d;\mathbb{R})$ we construct a Gaussian field whose sample-path Hölder exponent is $H(x) = m(x)-d/2$ at each point, and we prove that the existence threshold $m(x)>d/2$ is local: the regularity of the field near a point depends on $m$ only through its germ at that point. Finally we examine the canonical field, which supplies a direct check on both the threshold and the exponent. We use it to prove a rigidity statement: weighting the Matérn spectral density by $|k|^{-2}$ produces a superposition of ordinary Matérn fields of higher smoothness.

math.ST

A Novel Testing Approach for Differences Among Brain Connectomes

Statistical analysis on non-Euclidean spaces typically relies on distances as the primary tool for constructing likelihoods. However, manifold-valued data admits richer structures in addition to Riemannian distances. We demonstrate that simple, tractable models that do not rely exclusively on distances can be constructed on the manifold of symmetric positive definite (SPD) matrices, which naturally arises in brain connectivity analysis. Specifically, we highlight the manifold-valued Mahalanobis distribution, a parametric family that extends classical multivariate concepts to the SPD manifold. We develop estimators for this distribution and establish their asymptotic properties. Building on this framework, we propose a novel ANOVA test that leverages the manifold structure to obtain a test statistic that better captures the dimensionality of the data. We theoretically demonstrate that our test achieves superior statistical power compared to distance-based Fréchet ANOVA methods.

math.ST

riemtan, riemstats: R packages for Riemannian geometry techniques in the analysis of multiple samples of connectomes

Symmetric positive definite (SPD) matrices arising from functional connectivity analysis of neuroimaging data can be endowed with a Riemannian geometric structure that standard methods fail to respect. While existing R packages provide some tools for SPD matrix analysis, they suffer from limitations in scalability, numerical stability, and metric flexibility that hinder their application to modern large-scale connectomics studies. We present riemtan, a comprehensive R package that addresses these challenges through a unified, high-level interface supporting multiple Riemannian metrics, efficient parallel computation, and seamless conversion between manifold, tangent, and vectorized representations. Building on riemtan's foundation, we also introduce riemstats, which implements advanced statistical methods including Fréchet ANOVA, Riemannian ANOVA with classic test statistics, and harmonization techniques for multi-site studies. The modular design facilitates integration with existing R workflows and provides an extensible framework for future methodological developments in manifold-valued data analysis.

stat.CO

MECfda: An R Package for Bias Correction Due to Measurement Error in Functional and Scalar Covariates in Scalar-on-Function Regression Models

Functional data analysis (FDA) deals with high-resolution data recorded over a continuum, such as time, space or frequency. Device-based assessments of physical activity or sleep are objective yet still prone to measurement error. We present MECfda, an R package that (i) fits scalar-on-function, generalized scalar-on-function, and functional quantile regression models, and (ii) provides bias-corrected estimation when functional covariates are measured with error. By unifying these tools under a consistent syntax, MECfda enables robust inference for FDA applications that involve noisy functional data.

stat.ME