Abstract:Bayes classifiers for functional data pose a challenge. This is because probability density functions do not exist for functional data. As a consequence, the classical Bayes classifier using density quotients needs to be modified. We propose to use density ratios of projections on a sequence of eigenfunctions that are common to the groups to be classified. The density ratios can then be factored into density ratios of individual functional principal components whence the classification problem is reduced to a sequence of nonparametric one-dimensional density estimates. This is an extension to functional data of some of the very earliest nonparametric Bayes classifiers that were based on simple density ratios in the one-dimensional case. By means of the factorization of the density quotients the curse of dimensionality that would otherwise severely affect Bayes classifiers for functional data can be avoided. We demonstrate that in the case of Gaussian functional data, the proposed functional Bayes classifier reduces to a functional version of the classical quadratic discriminant. A study of the asymptotic behavior of the proposed classifiers in the large sample limit shows that under certain conditions the misclassification rate converges to zero, a phenomenon that has been referred to as "perfect classification". The proposed classifiers also perform favorably in finite sample applications, as we demonstrate in comparisons with other functional classifiers in simulations and various data applications, including wine spectral data, functional magnetic resonance imaging (fMRI) data for attention deficit hyperactivity disorder (ADHD) patients, and yeast gene expression data.

A Density-ratio Framework for Statistical Data Processing.

Probabilistic Data Fusion for Short-Term Traffic Prediction with Semiparametric Density Ratio Model

Density Estimation Based on Mass

Discriminative Density-ratio Estimation

Trimmed Density Ratio Estimation

A Statistical Information-Based Clustering Approach in Distance Space

Density-ratio Based Clustering for Discovering Clusters with Varying Densities.

Relative Density-Ratio Estimation for Robust Distribution Comparison

Density Ratio Estimation via Sampling along Generalized Geodesics on Statistical Manifolds

Demass: A New Density Estimator for Big Data

Out-of-Distribution Detection with Class Ratio Estimation

A density ratio framework for evaluating the utility of synthetic data

Semi-supervised learning with density-ratio estimation

Distributed Density Estimation Based on a Mixture of Factor Analyzers in a Sensor Network.

Estimating the Density Ratio between Distributions with High Discrepancy using Multinomial Logistic Regression

Telescoping Density-Ratio Estimation

A Unifying Framework for Some Directed Distances in Statistics

Optimal Bayes Classifiers for Functional Data and Density Ratios

A Distance Scaling Method to Improve Density-Based Clustering.

Density Ratio Peak Clustering

Density Estimation Over Data Stream ∗