Estimates of the approximation error using Rademacher complexity: Learning vector-valued functions.

Gnecco, Giorgio; Sanguineti, Marcello

Displaying similar documents to “Estimates of the approximation error using Rademacher complexity: Learning vector-valued functions.”

On approximation of continuous functions in the metric $\int_{a}^{b} | x (t) | d t$

Vlastimil Pták (1958)

Czechoslovak Mathematical Journal

Similarity:

An optimal error bound for a finite element approximation of a model for phase separation of a multi-component alloy with non-smooth free energy

John W. Barrett, James F. Blowey (1999)

ESAIM: Mathematical Modelling and Numerical Analysis - Modélisation Mathématique et Analyse Numérique

Similarity:

A note on the best $L_{2}$ approximation by ridge functions.

Ismailov, Vugar E. (2007)

Applied Mathematics E-Notes [electronic only]

Similarity:

On a theorem of V. Pták concerning best approximation of continuous functions in the metric $\int_{a}^{b} | x (t) | d t$

I. Singer (1960)

Czechoslovak Mathematical Journal

Similarity:

Multimodal discrete Karhunen-Loève expansion

Jiří Grim (1986)

Kybernetika

Similarity:

Integration with respect to a vector measure and function approximation.

Garcìa-Raffi, L.M., Ginestar, D., Sànchez-Pèrez, E.A. (2001)

Abstract and Applied Analysis

Similarity:

Best approximations, fixed points and parametric projections

Tiziana Cardinali (2002)

Discussiones Mathematicae, Differential Inclusions, Control and Optimization

Similarity:

If f is a continuous seminorm, we prove two f-best approximation theorems for functions Φ not necessarily continuous as a consequence of our version of Glebov's fixed point theorem. Moreover, we obtain another fixed point theorem that improves a recent result of [4]. In the last section, we study continuity-type properties of set valued parametric projections and our results improve recent theorems due to Mabizela [11].

Greedy Algorithms for Adaptive Approximation

Albert Cohen (2009)

Bollettino dell'Unione Matematica Italiana

Similarity:

We discuss the performances of greedy algorithms for two problems of numerical approximation. The first one is the best approximation of an arbitrary function by an N-terms linear combination of simple functions adaptively picked within a large dictionary. The second one is the approximation of an arbitrary function by a piecewise polynomial function on an optimally adapted triangulation of cardinality N. Performance is measured in terms of convergence rate with respect to the number...