Zobrazeno 1 - 10
of 196
pro vyhledávání: '"De Sterck Hans"'
Stochastic Reinforcement Learning with Stability Guarantees for Control of Unknown Nonlinear Systems
Designing a stabilizing controller for nonlinear systems is a challenging task, especially for high-dimensional problems with unknown dynamics. Traditional reinforcement learning algorithms applied to stabilization tasks tend to drive the system clos
Externí odkaz:
http://arxiv.org/abs/2409.08382
When applying nonnegative matrix factorization (NMF), generally the rank parameter is unknown. Such rank in NMF, called the nonnegative rank, is usually estimated heuristically since computing the exact value of it is NP-hard. In this work, we propos
Externí odkaz:
http://arxiv.org/abs/2407.00706
Graph Neural Networks (GNNs) have established themselves as the preferred methodology in a multitude of domains, ranging from computer vision to computational biology, especially in contexts where data inherently conform to graph structures. While ma
Externí odkaz:
http://arxiv.org/abs/2404.03081
Anderson acceleration (AA) is widely used for accelerating the convergence of an underlying fixed-point iteration $\bm{x}_{k+1} = \bm{q}( \bm{x}_{k} )$, $k = 0, 1, \ldots$, with $\bm{x}_k \in \mathbb{R}^n$, $\bm{q} \colon \mathbb{R}^n \to \mathbb{R}^
Externí odkaz:
http://arxiv.org/abs/2312.04776
Transformer-based models have achieved state-of-the-art performance in many areas. However, the quadratic complexity of self-attention with respect to the input length hinders the applicability of Transformer-based models to long sequences. To addres
Externí odkaz:
http://arxiv.org/abs/2310.11960
Publikováno v:
Journal of Space Weather and Space Climate, Vol 11, p 8 (2021)
The development of numerical models and tools which have operational space weather potential is an increasingly important area of research. This study presents recent Canadian efforts toward the development of a numerical framework for Sun-to-Earth s
Externí odkaz:
https://doaj.org/article/1e491443aec14797a925f1200e27b560
Autor:
Cui, Tiangang, De Sterck, Hans, Gilbert, Alexander D., Polishchuk, Stanislav, Scheichl, Robert
We develop new multilevel Monte Carlo (MLMC) methods to estimate the expectation of the smallest eigenvalue of a stochastic convection-diffusion operator with random coefficients. The MLMC method is based on a sequence of finite element (FE) discreti
Externí odkaz:
http://arxiv.org/abs/2303.03673
We study the combination of proximal gradient descent with multigrid for solving a class of possibly nonsmooth strongly convex optimization problems. We propose a multigrid proximal gradient method called MGProx, which accelerates the proximal gradie
Externí odkaz:
http://arxiv.org/abs/2302.04077
Training large neural networks is time consuming. To speed up the process, distributed training is often used. One of the largest bottlenecks in distributed training is communicating gradients across different nodes. Different gradient compression te
Externí odkaz:
http://arxiv.org/abs/2209.15203
Learning for control of dynamical systems with formal guarantees remains a challenging task. This paper proposes a learning framework to simultaneously stabilize an unknown nonlinear system with a neural controller and learn a neural Lyapunov functio
Externí odkaz:
http://arxiv.org/abs/2206.01913