Search
Results
-
Enhancing accuracy of deep learning algorithms by training with low-discrepancy sequences
(2020)SAM Research ReportWe propose a deep supervised learning algorithm based on low-discrepancy sequences as the training set. By a combination of theoretical arguments and extensive numerical experiments we demonstrate that the proposed algorithm significantly outperforms standard deep learning algorithms that are based on randomly chosen training data, for problems in moderately high dimensions. The proposed algorithm provides an efficient method for building ...Report -
Estimates on the generalization error of Physics Informed Neural Networks (PINNs) for approximating PDEs
(2020)SAM Research ReportPhysics informed neural networks (PINNs) have recently been widely used for robust and accurate approximation of PDEs. We provide rigorous upper bounds on the generalization error of PINNs approximating solutions of the forward problem for PDEs. An abstract formalism is introduced and stability properties of the underlying PDE are leveraged to derive an estimate for the generalization error in terms of the training error and number of ...Report -
Physics Informed Neural Networks for Simulating Radiative Transfer
(2020)SAM Research ReportWe propose a novel machine learning algorithm for simulating radiative transfer. Our algorithmis based on physics informed neural networks (PINNs), which are trained by minimizing the residualof the underlying radiative tranfer equations. We present extensive experiments and theoretical errorestimates to demonstrate that PINNs provide a very easy to implement, fast, robust and accuratemethod for simulating radiative transfer. We also ...Report -
Tunneling time and weak measurement in strong field ionisation
(2015)Research ReportTunnelling delays is a hotly debated topic, with many conflicting definitions and little consensus on when and if such definitions accurately describe the physical observables. Here we relate these different definitions to distinct experimental observables in strong field ionization, finding that two definitions, Larmor time and Bohmian time, are compatible with the attoclock observable and the resonance lifetime of a bound state, ...Report -
-
Well-balanced schemes for gravitationally stratified media
(2014)Research ReportWe present a well-balanced scheme for the Euler equations with gravitation. The scheme is capable of maintaining exactly (up to machine precision) a discrete hydrostatic equilibrium without any assumption on a thermodynamic variable such as specific entropy or temperature. The well-balanced scheme is based on a local hydrostatic pressure reconstruction. Moreover, it is computationally efficient and can be incorporated into any existing ...Report -
A machine learning framework for data driven acceleration of computations of differential equations
(2018)SAM Research ReportReport -
Gradient Gating for Deep Multi-Rate Learning on Graphs
(2022)SAM Research ReportWe present Gradient Gating (G2), a novel framework for improving the performance of Graph Neural Networks (GNNs). Our framework is based on gating the output of GNN layers with a mechanism for multi-rate flow of message passing information across nodes of the underlying graph. Local gradients are harnessed to further modulate message passing updates. Our framework flexibly allows one to use any basic GNN layer as a wrapper around which ...Report -
On the convergence of the spectral viscosity method for the incompressible Euler equations with rough initial data
(2019)SAM Research ReportReport