Designing universal causal deep learning models: The geometric (Hyper)transformer
From MaRDI portal
Publication:6196301
DOI10.1111/mafi.12389arXiv2201.13094OpenAlexW4367052271MaRDI QIDQ6196301
Gudmund Pammer, Anastasis Kratsios, Beatrice Acciaio
Publication date: 14 March 2024
Published in: Mathematical Finance (Search for Journal in Brave)
Full work available at URL: https://arxiv.org/abs/2201.13094
stochastic processesuniversal approximationmetric geometryrandom projectiongeometric deep learninghypernetworksadapted optimal transporttransformer networks
Artificial neural networks and deep learning (68T07) Optimal stochastic control (93E20) Financial applications of other theories (91G80)
Cites Work
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Unnamed Item
- Measure density and extension of Besov and Triebel-Lizorkin functions
- Optimal stochastic control, stochastic target problems, and backward SDE.
- Reservoir computing approaches to recurrent neural network training
- Theory of reproducing kernels and applications
- Adapted Wasserstein distances and stability in mathematical finance
- A functional extension of the Ito formula
- Information geometry and its applications
- Statistical analysis of counting processes
- Martingale Hardy spaces and their applications in Fourier analysis
- Exponential families of stochastic processes
- Multivariate normal distributions parametrized as a Riemannian symmetric space
- Lectures on analysis on metric spaces
- A random matrix approach to neural networks
- Wasserstein Riemannian geometry of Gaussian densities
- Provable approximation properties for deep neural networks
- The Cartan-Hadamard theorem for metric spaces with local geodesic bicombings
- On the computational power of neural nets
- Differential equations driven by rough signals. I: An extension of an inequality of L. C. Young
- Stochastic differential inclusions and applications.
- Functional Itō calculus and stochastic integral representation of martingales
- Approximation rates for neural networks with general activation functions
- Metric entropy limits on recurrent neural network learning of linear dynamical systems
- Echo state networks are universal
- All adapted topologies are equal
- A proof that rectified deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear heat equations
- Linear Lipschitz and \(C^1\) extension operators through random projection
- Information geometry and sufficient statistics
- Conical geodesic bicombings on subsets of normed vector spaces
- Convex geodesic bicombings and hyperbolicity
- On a Heath-Jarrow-Morton approach for stock options
- Bishop's generalized Stone-Weierstraß theorem for weighted spaces
- Measured descent: A new embedding method for finite metrics
- Neural networks in Fréchet spaces
- Barycenters in the Wasserstein Space
- Fixed point theorems for metric spaces with a conical geodesic bicombing
- The Wasserstein distance and approximation theorems
- Fuzzy Set Theory and Topos Theory
- Fading memory and the problem of approximating nonlinear operators with Volterra series
- Bond Pricing and the Term Structure of Interest Rates: A New Methodology for Contingent Claims Valuation
- One-Parameter Semigroups for Linear Evolution Equations
- Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations
- A phase transition phenomenon between the isometric and isomorphic extension problems for Hölder functions betweenLpspaces
- Uniform decomposition of probability measures: quantization, clustering and rate of convergence
- Convergence rate of optimal quantization grids and application to empirical measure
- Deep Network With Approximation Error Being Reciprocal of Width to Power of Square Root of Depth
- Equal risk pricing of derivatives with deep hedging
- Solving high-dimensional optimal stopping problems using deep learning
- Memory Capacity of Neural Networks with Threshold and Rectified Linear Unit Activations
- Randomized Wasserstein Barycenter Computation: Resampling with Statistical Guarantees
- Quant GANs: deep generation of financial time series
- Stability and memory-loss go hand-in-hand: three results in dynamics and computation
- Deep hedging
- Universal features of price formation in financial markets: perspectives from deep learning
- DeepLOB: Deep Convolutional Neural Networks for Limit Order Books
- Stochastic Calculus and Applications
- Echo State Property Linked to an Input: Exploring a Fundamental Characteristic of Recurrent Neural Networks
- Information Geometry
- Stochastic Gradient Descent on Riemannian Manifolds
- Learning representations by back-propagating errors
- Deep optimal stopping
- A generalized modulus of smoothness
- HJM: A Unified Approach to Dynamic Models for Fixed Income, Credit and Equity Markets
- Book Review: Optimal transport: old and new
- Theory of Reproducing Kernels
- A logical calculus of the ideas immanent in nervous activity
- Consistency problems for Heath-Jarrow-Morton interest rate models
- Bilipschitz embeddings of metric spaces into space forms
- Neural network approximation: three hidden layers are enough
- Fading memory echo state networks are universal
