Search

We propose a sparse algebra for samplet compressed kernel matrices, to enable efficient scattered data analysis. We show the compression of kernel matrices by means of samplets produces optimally sparse matrices in a certain S-format. It can be performed in cost and memory that scale essentially linearly with the matrix size $N$, for kernels of finite differentiability, along with addition and multiplication of S-formatted matrices. We prove and exploit the fact that the inverse of a kernel matrix (if it exists) is compressible in the S-format as well. Selected inversion allows to directly compute the entries in the corresponding sparsity pattern. The S-formatted matrix operations enable the efficient, approximate computation of more complicated matrix functions such as ${\bm A}^\alpha$ or $\exp({\bm A})$. The matrix algebra is justified mathematically by pseudo differential calculus. As an application, efficient Gaussian process learning algorithms for spatial statistics is considered. Numerical results are presented to illustrate and quantify our findings.

Homogenization in terms of multiscale limits transforms a multiscale problem with $n+1$ asymptotically separated microscales posed on a physical domain $D \subset \mathbb{R}^d$ into a one-scale problem posed on a product domain of dimension $(n+1)d$ by introducing $n$ so-called "fast variables". This procedure allows to convert $n+1$ scales in $d$ physical dimensions into a single-scale structure in $(n+1)d$ dimensions. We prove here that both the original, physical multiscale problem and the corresponding high-dimensional, one-scale limiting problem can be efficiently treated numerically with the recently developed quantized tensor-train finite-element method (QTT-FEM). The method is based on restricting computation to sequences of nested subspaces of low dimensions (which are called tensor ranks) within a vast but generic "virtual" (background) discretization space. In the course of computation, these subspaces are computed iteratively and data-adaptively at runtime, bypassing any "offline precomputation". For the purpose of theoretical analysis, such low-dimensional subspaces are constructed analytically to bound the tensor ranks vs. error $\tau>0$. We consider a model linear elliptic multiscale problem in several physical dimensions and show, theoretically and experimentally, that both (i) the solution of the associated high-dimensional one-scale problem and (ii) the corresponding approximation to the solution of the multiscale problem admit efficient approximation by the QTT-FEM. These problems can therefore be numerically solved in a scale-robust fashion by standard (low-order) PDE discretizations combined with state-of-the-art general-purpose solvers for tensor-structured linear systems. We prove scale-robust exponential convergence, i.e., that QTT-FEM achieves accuracy $\tau$ with the number of effective degrees of freedom scaling polynomially in $\log \tau$., Comment: 31 pages, 8 figures

We present a novel algorithmic approach and an error analysis leveraging Quasi-Monte Carlo points for training deep neural network (DNN) surrogates of Data-to-Observable (DtO) maps in engineering design. Our analysis reveals higher-order consistent, deterministic choices of training points in the input data space for deep and shallow Neural Networks with holomorphic activation functions such as tanh. These novel training points are proved to facilitate higher-order decay (in terms of the number of training samples) of the underlying generalization error, with consistency error bounds that are free from the curse of dimensionality in the input data space, provided that DNN weights in hidden layers satisfy certain summability conditions. We present numerical experiments for DtO maps from elliptic and parabolic PDEs with uncertain inputs that confirm the theoretical analysis.

We define the higher order moments associated to the stochastic solution of an elliptic BVP in D⊂ℝ d with stochastic input data. We prove that the k-th moment solves a deterministic problem in D k ⊂ℝ dk , for which we discuss well-posedness and regularity. We discretize the deterministic k-th moment problem using sparse grids and, exploiting a spline wavelet basis, we propose an efficient algorithm, of logarithmic-linear complexity, for solving the resulting system

This paper is a contemporary review of quasi-Monte Carlo (QMC) methods, that is, equal-weight rules for the approximate evaluation of high-dimensional integrals over the unit cube [0,1]s. It first introduces the by-now standard setting of weighted Hilbert spaces of functions with square-integrable mixed first derivatives, and then indicates alternative settings, such as non-Hilbert spaces, that can sometimes be more suitable. Original contributions include the extension of the fast component-by-component (CBC) construction of lattice rules that achieve the optimal convergence order (a rate of almost 1/N, where N is the number of points, independently of dimension) to so-called "product and order dependent” (POD) weights, as seen in some recent applications. Although the paper has a strong focus on lattice rules, the function space settings are applicable to all QMC methods. Furthermore, the error analysis and construction of lattice rules can be adapted to polynomial lattice rules from the family of digital nets

The problem of a-posteriori estimation of the modelling error and the adaptive modelling of plates is addressed in this paper. A family of models is proposed and upper and lower a-posteriori estimates of the modelling error in various norms are provided. The estimates are locally asymptotically exact.

Le modèle développé à l'Institut universitaire de médecine sociale et préventive de Lausanne utilise un programme informatique pour simuler les mouvements d'entrées et de sorties des hôpitaux de soins généraux. Cette simulation se fonde sur les données récoltées de routine dans les hôpitaux; elle tient notamment compte de certaines variations journalières et saisonnières, du nombre d'entrées, ainsi que du "Case-Mix" de l'hôpital, c'est-à-dire de la répartition des cas selon les groupes cliniques et l'âge des patients.

Le modèle développé à l'Institut universitaire de médecine sociale et préventive de Lausanne utilise un programme informatique pour simuler les mouvements d'entrées et de sorties des hôpitaux de soins généraux. Cette simulation se fonde sur les données récoltées de routine dans les hôpitaux; elle tient notamment compte de certaines variations journalières et saisonnières, du nombre d'entrées, ainsi que du "Case-Mix" de l'hôpital, c'est-à-dire de la répartition des cas selon les groupes cliniques et l'âge des patients.

La projection utilise toujours le programme de simulation SIMULIT, dans sa treizième version. (...) Seule l'évolution démographique a été considérée dans les projections du nombre de lits: aucune des autres variables susceptibles de changer dans le futur n'a été prise en compte, ni celle en relation avec l'activité hospitalière elle-même (modification des taux d'hospitalisation, des durées de séjour, etc.), ni celles concernant l'état de santé de la population (modification de l'incidence ou de la prévalence des maladies). En d'autres termes, cette projection montre l'effet de l'évolution démographique sur l'activité hospitalière, si les caractéristiques de cette activité devaient rester celles observées dans les années 80. Il ne s'agit donc pas d'une prévision. [Auteurs, p. 1]

We define the higher order moments associated to the stochastic solution of an elliptic BVP in D⊂ℝ d with stochastic input data. We prove that the k-th moment solves a deterministic problem in D k ⊂ℝ dk , for which we discuss well-posedness and regularity. We discretize the deterministic k-th moment problem using sparse grids and, exploiting a spline wavelet basis, we propose an efficient algorithm, of logarithmic-linear complexity, for solving the resulting system

This paper is a contemporary review of quasi-Monte Carlo (QMC) methods, that is, equal-weight rules for the approximate evaluation of high-dimensional integrals over the unit cube [0,1]s. It first introduces the by-now standard setting of weighted Hilbert spaces of functions with square-integrable mixed first derivatives, and then indicates alternative settings, such as non-Hilbert spaces, that can sometimes be more suitable. Original contributions include the extension of the fast component-by-component (CBC) construction of lattice rules that achieve the optimal convergence order (a rate of almost 1/N, where N is the number of points, independently of dimension) to so-called "product and order dependent” (POD) weights, as seen in some recent applications. Although the paper has a strong focus on lattice rules, the function space settings are applicable to all QMC methods. Furthermore, the error analysis and construction of lattice rules can be adapted to polynomial lattice rules from the family of digital nets