A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Study on solutions of degenerate equations on manifolds, linking behavior to geometry and decay rates.
DistillKac generates images quickly using damped wave equations.
We introduce economic models based on Boolean Delay Equations: this formalism makes easier to take into account the complexity of the interactions between firms and is particularly appropriate for studying the propagation of an initial damage due to a catastrophe. Here we concentrate on simple cases, which allow to und…
We study the problem of coupling Einstein's equations to a relativistic and physically well-motivated version of the Navier-Stokes equations. Under a natural evolution condition for the vorticity, we prove existence and uniqueness in a suitable Gevrey class if the fluid is incompressible, where this condition is given …
We investigate shock-wave solutions of the Einstein equations in the case when the speed of propagation is equal to the speed of light. The work extends the shock matching theory of Smoller and Temple, which characterizes solutions of the Einstein equations when the spacetime metric is only Lipschitz continuous across …
Infinite-time blow-up in high-dimensional mean curvature flow.
Uniform-in-time analysis for Stein Variational Gradient Descent across various metrics.
Using a simple and well-motivated modification of the stress-energy tensor for a viscous fluid proposed by Lichnerowicz, we prove that Einstein's equations coupled to a relativistic version of the Navier-Stokes equations are well-posed in a suitable Gevrey class if the fluid is incompressible and irrotational. These la…
We investigate the initial value problem for the Einstein-Euler equations of general relativity under the assumption of Gowdy symmetry on T3, and we construct matter spacetimes with low regularity. These spacetimes admit, both, impulsive gravitational waves in the metric (for instance, Dirac mass curvature singularitie…
Study shows critical initialisation not crucial for ReLU networks under dropout limits.
SingleProp speeds up robust neural network training with minimal certification.
PGMax automates PGM inference on GPUs, improving quality and speed.
Field theory explains optimal scaling in ResNets for signal propagation.
Deep neural network reconstructs traffic speeds from sparse vehicle data.
The Resilient Propagation (Rprop) algorithm has been very popular for backpropagation training of multilayer feed-forward neural networks in various applications. The standard Rprop however encounters difficulties in the context of deep neural networks as typically happens with gradient-based learning algorithms. In th…
Deep networks with orthogonal weights show stable fluctuations, improving generalization and training speed.
Finite-time extinction and smoothing effects in fractional fast diffusion on manifolds.
FiniteNet uses a neural network to improve PDE solving methods.
Study the geometry of gas giant planets to infer their internal structure.
EP method speeds up Bayesian probit regression in high dimensions.
A new stochastic algorithm approximates optimal distributions without requiring propagation of chaos.
A new interpolation method speeds up neural ODE training.
Deep vanilla transformers trained without shortcuts achieve similar performance to standard models.
Deep learning speeds up material property quantification using stress waves.
Researchers reconstruct simple Riemannian manifolds from boundary wave arrival times.
Loopy and generalized belief propagation are popular algorithms for approximate inference in Markov random fields and Bayesian networks. Fixed points of these algorithms correspond to extrema of the Bethe and Kikuchi free energy. However, belief propagation does not always converge, which explains the need for approach…
Modeling of a wide class of physical phenomena, such as crystal growth and flame propagation, leads to tracking fronts moving with curvature-dependent speed. When the speed is the curvature this leads to one of the classical degenerate nonlinear second order differential equations on Euclidean space. One naturally wond…
Deep neural networks have gained tremendous popularity in last few years. They have been applied for the task of classification in almost every domain. Despite the success, deep networks can be incredibly slow to train for even moderate sized models on sufficiently large datasets. Additionally, these networks require l…
We give a complete proof of a propagation theorem of multiplicity-free property from fibers to spaces of global sections for holomorphic vector bundles. The propagation theorem is formalised in three ways, aiming for producing various multiplicity-free theorems in representation theory for both finite and infinite dime…
Many inference problems involving questions of optimality ask for the maximum or the minimum of a finite set of unknown quantities. This technical report derives the first two posterior moments of the maximum of two correlated Gaussian variables and the first two posterior moments of the two generating variables (corre…
Classifies self-similar solutions for heat equations with positive speed.
The purpose of this paper is to advance the understanding of the conditions that give rise to flash crash contagion, particularly with respect to overlapping asset portfolio crowding. To this end, we designed, implemented, and assessed a hybrid micro-macro agent-based model, where price impact arises endogenously throu…
Despite the widespread practical success of deep learning methods, our theoretical understanding of the dynamics of learning in deep neural networks remains quite sparse. We attempt to bridge the gap between the theory and practice of deep learning by systematically analyzing learning dynamics for the restricted case o…
Paper compares hard and soft EM for BN learning from incomplete data.
Study traveling waves in hyperbolic space for Fisher-KPP equations.
Gaussian processes (GP) are attractive building blocks for many probabilistic models. Their drawbacks, however, are the rapidly increasing inference time and memory requirement alongside increasing data. The problem can be alleviated with compactly supported (CS) covariance functions, which produce sparse covariance ma…
An analytico-geometric reflection principle is established by means of normal deformations of analytic discs.
New EP variants improve inference stability and efficiency.
Study of fastest paths in anisotropic media via Finsler geometry.
Humans gain an implicit understanding of physical laws through observing and interacting with the world. Endowing an autonomous agent with an understanding of physical laws through experience and observation is seldom practical: we should seek alternatives. Fortunately, many of the laws of behaviour of the physical wor…
A new method prunes activation gradients to speed up CNN training.
The perturbative Chern-Simons theory is studied in a finite-dimensional version or assuming that the propagator satisfies certain properties (as is the case, e.g., with the propagator defined by Axelrod and Singer). It turns out that the effective BV action is a function on cohomology (with shifted degrees) that solves…
Efficiently verifies neural networks by handling neuron splits, improving speed and accuracy.
Proposes a label propagation framework for domain adaptation.
In this paper, we will show an unprecedented method to accelerate training and improve performance, which called random gradient (RG). This method can be easier to the training of any model without extra calculation cost, we use Image classification, Semantic segmentation, and GANs to confirm this method can improve sp…
Study reconstructs Riemannian metric from Cherenkov radiation in complex media.
SparseTrain uses dynamic sparsity in training deep neural networks on CPUs.
New methods speed up training of differentially private deep learning models.