Abstract:We introduce the concept of mixture inner product spaces associated with a given separable Hilbert space, which feature an infinite-dimensional mixture of finite-dimensional vector spaces and are dense in the underlying Hilbert space. Any Hilbert valued random element can be arbitrarily closely approximated by mixture inner product space valued random elements. While this concept can be applied to data in any infinite-dimensional Hilbert space, the case of functional data that are random elements in the L space of square integrable functions is of special interest. For functional data, mixture inner product spaces provide a new perspective, where each realization of the underlying stochastic process falls into one of the component spaces and is represented by a finite number of basis functions, the number of which corresponds to the dimension of the component space. In the mixture representation of functional data, the number of included mixture components used to represent a given random element in L is specifically adapted to each random trajectory and may be arbitrarily large. Key benefits of this novel approach are, first, that it provides a new perspective on the construction of a probability density in function space under mild regularity conditions, and second, that individual trajectories possess a trajectory-specific dimension that corresponds to a latent random variable, making it possible to use a larger number of components for less smooth and a smaller number for smoother trajectories. This enables flexible and parsimonious modeling of heterogeneous trajectory shapes. We establish estimation consistency of the functional mixture density and introduce an algorithm for fitting the functional mixture model based on a modified expectation-maximization algorithm. Simulations confirm that in comparison to traditional functional principal component analysis the proposed method achieves similar or better data recovery while using fewer components on average. Its practical merits are also demonstrated in an analysis of egg-laying trajectories for medflies.

Random Locations, Ordered Random Sets and Stationarity

On Global and Local Properties of the Trajectories of Gaussian Random Fields—A Look Through the Set of Limit Points

The moduli of non-differentiability for Gaussian random fields with stationary increments

Nonparametric M-estimation for Functional Stationary Ergodic Data

Random Fields in Space and Time

Iterated Piecewise-Stationary Random Functions

Randomized Strategic Facility Location with Predictions

Limit Theorems in the Nonparametric Conditional Single-Index U-Processes for Locally Stationary Functional Random Fields under Stochastic Sampling Design

Random Marked Sets

A family of random sup-measures with long-range dependence

Functional linear regression with points of impact

Proximal random reshuffling under local Lipschitz continuity

On the Quenched Functional Central Limit Theorem for Stationary Random Fields under Projective Criteria

A Class of Stationary Random Fields with a Simple Correlation Structure

Local stationarity and time-inhomogeneous Markov chains

Finitude of physical measures for random maps

One-to-one correspondences between discrete multivariate stationary, self-similar and stationary increment fields

Roduct s paces and t heir a pplication to f unctional d ata a nalysis

Small values and functional laws of the iterated logarithm for operator fractional Brownian motion

Critical Intensity for Unbounded Sequential Localizability.

Analysis of Random Walk Mobility Models with Location Heterogeneity