Projects per year
Abstract
We study two-layer neural networks whose domain and range are Banach spaces with separable preduals. In addition, we assume that the image space is equipped with a partial order, i.e. it is a Riesz space. As the nonlinearity we choose the lattice operation of taking the positive part; in case of $\mathbb R^d$-valued neural networks this corresponds to the ReLU activation function. We prove inverse and direct approximation theorems with Monte-Carlo rates for a certain class of functions, extending existing results for the finite-dimensional case. In the second part of the paper, we study, from the regularisation theory viewpoint, the problem of finding optimal representations of such functions via signed measures on a latent space from a finite number of noisy observations. We discuss regularity conditions known as source conditions and obtain convergence rates in a Bregman distance for the representing measure in the regime when both the noise level goes to zero and the number of samples goes to infinity at appropriate rates.
Original language | English |
---|---|
Pages (from-to) | 6358-6389 |
Number of pages | 32 |
Journal | SIAM Journal on Mathematical Analysis |
Volume | 54 |
Issue number | 6 |
Early online date | 8 Dec 2022 |
DOIs | |
Publication status | Published - 31 Dec 2022 |
Bibliographical note
Funding:The work of the author was supported by EPSRC Fellowship grant EP/V003615/1, the Cantab Capital Institute for the Mathematics of Information at the University of Cambridge, and the National Physical Laboratory
Keywords
- cs.LG
- cs.NA
- math.FA
- math.NA
- math.PR
- 68Q32, 68T07, 46E40, 41A65, 65J22
Fingerprint
Dive into the research topics of 'Two-layer neural networks with values in a Banach space'. Together they form a unique fingerprint.Projects
- 1 Finished
-
Regularisation theory in the data driven setting
Korolev, Y. (PI)
Engineering and Physical Sciences Research Council
1/09/22 → 31/10/24
Project: Research council