EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Designing universal causal deep learning models: The geometric (Hyper)transformer

Beatrice Acciaio, Anastasis Kratsios, Gudmund Pammer

Mathematical Finance · 2024-04 · 34 (2) · 671-735

DOI: 10.1111/mafi.12389 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Mathematical Finance, online erschienen am 26.04.2023. Verfasst von Beatrice Acciaio und 2 weiteren Autor:innen. Die Arbeit wurde bisher 15-mal zitiert (Crossref) und nennt 169 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: H2020 European Research Council, ETH Zürich Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Several problems in stochastic analysis are defined through their geometry, and preserving that geometric structure is essential to generating meaningful predictions. Nevertheless, how to design principled deep learning (DL) models capable of encoding these geometric structures remains largely unknown. We address this open problem by introducing a universal causal geometric DL framework in which the user specifies a suitable pair of metric spaces and and our framework returns a DL model capable of causally approximating any “regular” map sending time series in to time series in while respecting their forward flow of information throughout time. Suitable geometries on include various (adapted) Wasserstein spaces arising in optimal stopping problems, a variety of statistical manifolds describing the conditional distribution of continuous‐time finite state Markov chains, and all Fréchet spaces admitting a Schauder basis, for example, as in classical finance. Suitable spaces are compact subsets of any Euclidean space. Our results all quantitatively express the number of parameters needed for our DL model to achieve a given approximation error as a function of the target map's regularity and the geometric structure both of and of . Even when omitting any temporal structure, our universal approximation theorems are the first guarantees that Hölder functions, defined between such and can be approximated by DL models.

Förderung

  • H2020 European Research Council
  • ETH Zürich Foundation

Zitieren

APA

Acciaio, B., Kratsios, A., & Pammer, G. (2024). Designing universal causal deep learning models: The geometric (Hyper)transformer. Mathematical Finance, 34(2), 671-735. https://doi.org/10.1111/mafi.12389

Vancouver

Acciaio B, Kratsios A, Pammer G. Designing universal causal deep learning models: The geometric (Hyper)transformer. Mathematical Finance. 2024;34(2):671-735. doi:10.1111/mafi.12389

BibTeX
@article{Acciaio2024,
  author = {Acciaio, Beatrice and Kratsios, Anastasis and Pammer, Gudmund},
  title = {Designing universal causal deep learning models: The geometric (Hyper)transformer},
  journal = {Mathematical Finance},
  year = {2024},
  volume = {34},
  number = {2},
  pages = {671--735},
  doi = {10.1111/mafi.12389}
}
Kostenlos starten →