Ch. Schwab

H. Harbrecht, M. Multerer, O. Schenk et al.

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 ${\bm A}^α$ 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.

F. Rohner, Ch. Schwab, C. Xenophontos

We study finite-element and deep feedforward neural network (DNN for short) expressivity rate bounds for solution sets of a model linear, second order singularly perturbed, elliptic two-point boundary value problem, in Sobolev norms on a bounded interval $(-1,1)$, with explicit dependence on the singular perturbation parameter $\e\in (0,1]$. Emphasis is on low Sobolev regularity of the data, i.e., source term $f$ and reaction coefficient $b$. A proof of $\e$-explicit solution regularity based on exponentially weighted energy-norm bounds is developed, and \emph{$\e$-robust, algebraic expression rate bounds} in Sobolev norms for $\mathbb{P}_1$ Finite-Elements on exponential and Shishkin type meshes is proved. Expression rates for shallow (fixed depth) $\ReLU$-NNs are shown which are robust w.r. to $\e$ and explicit in terms of the NN size. Robust NN expression rate bounds are further studied for deep feedforward DNNs with ReLU and tanh-activations. As in \cite{OSX24_1085}, tanh- and sigmoid-activated sub-NNs allow to include exponential boundary layer functions exactly into the NN feature space, leading to reduced NN sizes. Recent bitstring encoding techniques for deep NNs with ReLU activations afford, still under low data regularity $f,b \in H^1(I)$ \emph{twice the (robust) convergence rate of $\mathbb{P}_1$ Finite-Elements} achievable with ``eXp'' or Shishkin meshes.

M. Longo, S. Mishra, T. K. Rusch et al.

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.