Abstract:The challenge of approximating functions in infinite-dimensional spaces from finite samples is widely regarded as formidable. We delve into the challenging problem of the numerical approximation of Sobolev-smooth functions defined on probability spaces. Our particular focus centers on the Wasserstein distance function, which serves as a relevant example. In contrast to the existing body of literature focused on approximating efficiently pointwise evaluations, we chart a new course to define functional approximants by adopting three machine learning-based approaches: 1. Solving a finite number of optimal transport problems and computing the corresponding Wasserstein potentials. 2. Employing empirical risk minimization with Tikhonov regularization in Wasserstein Sobolev spaces. 3. Addressing the problem through the saddle point formulation that characterizes the weak form of the Tikhonov functional's Euler-Lagrange equation. We furnish explicit and quantitative bounds on generalization errors for each of these solutions. We leverage the theory of metric Sobolev spaces and we combine it with techniques of optimal transport, variational calculus, and large deviation bounds. In our numerical implementation, we harness appropriately designed neural networks to serve as basis functions. These networks undergo training using diverse methodologies. This approach allows us to obtain approximating functions that can be rapidly evaluated after training. Our constructive solutions significantly enhance at equal accuracy the evaluation speed, surpassing that of state-of-the-art methods by several orders of magnitude. This allows evaluations over large datasets several times faster, including training, than traditional optimal transport algorithms. Our analytically designed deep learning architecture slightly outperforms the test error of state-of-the-art CNN architectures on datasets of images.

Approximation by filter functions

General Approach to Function Approximation

Approximating Markov Processes Through Filtration

Almost-sure quasi-optimal approximation in reproducing kernel Hilbert spaces

An Approximation Theory Framework for Measure-Transport Sampling Algorithms

Density in Approximation Theory

Approximation in Hilbert spaces of the Gaussian and related analytic kernels

On the approximation of vector-valued functions by volume sampling

One Sample Fits All: Approximating All Probabilistic Values Simultaneously and Efficiently

Approximate Bayesian Computation via Classification

Best Simultaneous Approximation of an Infinite Set of Functions

Approximation from Noisy Data

Approximation Theory, Computing, and Deep Learning on the Wasserstein Space

Approximate filtering via discrete dual processes

Adaptive tempering schedules with approximative intermediate measures for filtering problems

On the power of function values for the approximation problem in various settings

Statistical Accuracy of Approximate Filtering Methods

Approximation properties of certain operator-induced norms on Hilbert spaces

On the Approximation of Kernel functions

An Approximation Theory for Metric Space-Valued Functions With A View Towards Deep Learning

Approximation by Finite Supported Functions