Functional data analysis for density functions by transformation to a Hilbert space

DOI10.1214/15-AOS1363zbMATH Open1331.62203arXiv1601.02869MaRDI QIDQ5963520FDOQ5963520

Authors: Alexander Petersen, Hans-Georg Müller 

Publication date: 22 February 2016

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: Functional data that are nonnegative and have a constrained integral can be considered as samples of one-dimensional density functions. Such data are ubiquitous. Due to the inherent constraints, densities do not live in a vector space and, therefore, commonly used Hilbert space based methods of functional data analysis are not applicable. To address this problem, we introduce a transformation approach, mapping probability densities to a Hilbert space of functions through a continuous and invertible map. Basic methods of functional data analysis, such as the construction of functional modes of variation, functional regression or classification, are then implemented by using representations of the densities in this linear space. Representations of the densities themselves are obtained by applying the inverse map from the linear functional space to the density space. Transformations of interest include log quantile density and log hazard transformations, among others. Rates of convergence are derived for the representations that are obtained for a general class of transformations under certain structural properties. If the subject-specific densities need to be estimated from data, these rates correspond to the optimal rates of convergence for density estimation. The proposed methods are illustrated through simulations and applications in brain imaging.

Full work available at URL: https://arxiv.org/abs/1601.02869

Recommendations

• Preprocessing of centred logratio transformed density functions using smoothing splines
• Nonparametric density estimation for functional data via wavelets
• Hilbert space of probability density functions based on Aitchison geometry
• Dimensionality reduction when data are density functions
• Compositional splines for representation of density functions

zbMATH Keywords

predictionkernel estimationrate of convergencequantilesWasserstein metricbasis representationlog hazardsamples of density functions

Mathematics Subject Classification ID

Nonparametric estimation (62G05) Density estimation (62G07) Asymptotic properties of nonparametric inference (62G20)

Cites Work

• Simplicial principal component analysis for density functions in Bayes spaces
• Beta kernel estimators for density functions
• Functional data analysis.
• Remarks on a Multivariate Transformation
• Linear processes in function spaces. Theory and applications
• Title not available (Why is that?)
• Consistency of the beta kernel density function estimator
• Robust functional principal components: a projection-pursuit approach
• Functional additive models
• On Properties of Functional Principal Components Analysis
• Introduction to nonparametric estimation
• Asymptotic theory for the principal component analysis of a vector random function: Some applications to statistical inference
• Principal components analysis of sampled functions
• Prediction in functional linear regression
• Properties of principal component methods for functional and longitudinal data analysis
• Hilbert space of probability density functions based on Aitchison geometry
• Inference for Density Families Using Functional Principal Component Analysis
• Dimensionality reduction when data are density functions
• Uniform convergence rates for nonparametric regression and principal component analysis in functional/longitudinal data
• Functional density synchronization
• Asymptotic Minimax Character of the Sample Distribution Function and of the Classical Multinomial Estimator
• Title not available (Why is that?)
• WHICH PART OF THE SAMPLE CONTAINS THE INFORMATION?
• Title not available (Why is that?)
• Transformations in Density Estimation
• Nonparametric Statistical Data Modeling
• Common functional principal components
• Title not available (Why is that?)
• Methodology and convergence rates for functional linear regression
• Principal Modes of Variation for Processes with Continuous Sample Curves
• A note on the shape of the generalizedc-numerical range
• A Note on Asymptotic Joint Normality
• On improving density estimators which are not bona fide functions
• Dynamical Correlation for Multivariate Longitudinal Data
• Multivariate Boundary Kernels and a Continuous Least Squares Principle
• Estimating densities, quantiles, quantile densities and density quantiles

Cited In (52)

• Detection and localization of changes in a panel of densities
• Wasserstein Regression
• Cointegrated linear processes in Bayes Hilbert space
• On the Schoenberg transformations in data analysis: theory and illustrations
• Describing the concentration of income populations by functional principal component analysis on Lorenz curves
• Testing linear operator constraints in functional response regression with incomplete response functions
• Additive Functional Regression for Densities as Responses
• Spherical autoregressive models, with application to distributional and compositional time series
• Nonparametric Estimation of Repeated Densities with Heterogeneous Sample Sizes
• Wasserstein \(F\)-tests and confidence bands for the Fréchet regression of density response curves
• Principal component analysis for functional data on Riemannian manifolds and spheres
• Multi-feature clustering of step data using multivariate functional principal component analysis
• Hypothesis Testing for Matched Pairs with Missing Data by Maximum Mean Discrepancy: An Application to Continuous Glucose Monitoring
• Variable selection for functional additive models and an application to the population age structure data
• Mixture inner product spaces and their application to functional data analysis
• Shape-preserving prediction for stationary functional time series
• Wasserstein gradients for the temporal evolution of probability distributions
• Latent deformation models for multivariate functional data and time-warping separability
• Distribution-on-distribution regression with Wasserstein metric: multivariate Gaussian case
• Direct Bayesian linear regression for distribution-valued covariates
• Functional data analysis in the Banach space of continuous functions
• Hilbert space of probability density functions based on Aitchison geometry
• Statistical inference on the Hilbert sphere with application to random densities
• Locally polynomial Hilbertian additive regression
• On distributional autoregression and iterated transportation
• Asynchronous Functional Linear Regression Models for Longitudinal Data in Reproducing Kernel Hilbert Space
• Subgroup analysis for functional partial linear regression model
• INFERENCE ON THE DIMENSION OF THE NONSTATIONARY SUBSPACE IN FUNCTIONAL TIME SERIES
• Monitoring for a change point in a sequence of distributions
• Bivariate densities in Bayes spaces: orthogonal decomposition and spline representation
• Mean and covariance estimation for discretely observed high-dimensional functional data: rates of convergence and division of observational regimes
• Continuous and atlas-free analysis of brain structural connectivity
• Hilbertian additive regression with parametric help
• Upper and lower risk bounds for estimating the Wasserstein barycenter of random measures on the real line
• Additive regression for predictors of various natures and possibly incomplete Hilbertian responses
• Concurrent object regression
• Additive regression with Hilbertian responses
• Functional Outlier Detection for Density-Valued Data with Application to Robustify Distribution-to-Distribution Regression
• COINTEGRATION IN FUNCTIONAL AUTOREGRESSIVE PROCESSES
• Dimension Reduction for Fréchet Regression
• Multivariate scalar on multidimensional distribution regression with application to modeling the association between physical activity and cognitive functions
• Wasserstein autoregressive models for density time series
• Geodesic PCA versus Log-PCA of Histograms in the Wasserstein Space
• Point process models for COVID-19 cases and deaths
• Robust functional principal component analysis for non-Gaussian longitudinal data
• Density estimation with distribution element trees
• A general framework for multivariate functional principal component analysis of amplitude and phase variation
• Weighting the domain of probability densities in functional data analysis
• Regression Analysis of Asynchronous Longitudinal Functional and Scalar Data
• M-LDQ feature embedding and regression modeling for distribution-valued data
• Machine learning and optimal transport: some statistical and algorithmic tools
• Metric statistics: exploration and inference for random objects with distance profiles

Uses Software

• fda (R)

This page was built for publication: Functional data analysis for density functions by transformation to a Hilbert space

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5963520)