Abstract:We introduce a new framework to analyze shape descriptors that capture the geometric features of an ensemble of point clouds. At the core of our approach is the point of view that the data arises as sampled recordings from a metric space-valued stochastic process, possibly of nonstationary nature, thereby integrating geometric data analysis into the realm of functional time series analysis. Our framework allows for natural incorporation of spatial-temporal dynamics, heterogeneous sampling, and the study of convergence rates. Further, we derive complete invariants for classes of metric space-valued stochastic processes in the spirit of Gromov, and relate these invariants to so-called ball volume processes. Under mild dependence conditions, a weak invariance principle in $D([0,1]\times [0,\mathscr{R}])$ is established for sequential empirical versions of the latter, assuming the probabilistic structure possibly changes over time. Finally, we use this result to introduce novel test statistics for topological change, which are distribution-free in the limit under the hypothesis of stationarity. We explore these test statistics on time series of single-cell mRNA expression data, using shape descriptors coming from topological data analysis.

Foundations of Structural Statistics: Statistical Manifolds

Foundations of Structural Statistics: Topological Statistical Theory

Applications of Structural Statistics: Geometric Inference in Exponential Families

Riemannian Statistics for Any Type of Data

Geometric Structures Induced by Deformations of the Legendre Transform

Locally conformally Hessian and statistical manifolds

A statistical framework for analyzing shape in a time series of random geometric objects

Statistical Methods in Topological Data Analysis for Complex, High-Dimensional Data

Riemannian geometry for shape analysis and computational anatomy

On manifold-like polyfolds as differential geometrical objects with applications in complex geometry

Differential Geometry for Model Independent Analysis of Images and Other Non-Euclidean Data: Recent Developments

On a Torsion/Curvature Analogue of Dual Connections and Statistical Manifolds

Statistics on the Manifold of Multivariate Normal Distributions: Theory and Application to Diffusion Tensor MRI Processing

Diffusion Geometry

Intrinsic Finite Element Error Analysis on Manifolds with Regge Metrics, with Applications to Calculating Connection Forms

Configurational microcanonical statistical mechanics from Riemannian geometry of equipotenital level sets

Generalized Differential Geometry

Equiaffine Structure and Conjugate Ricci-symmetry of a Statistical Manifold

Building Three-Dimensional Differentiable Manifolds Numerically II: Limitations

Manifold Diffusion Geometry: Curvature, Tangent Spaces, and Dimension

Affine statistical bundle modeled on a Gaussian Orlicz-Sobolev space