Classically transmission conditions between subdomains are optimized for a simplified two subdomain decomposition to obtain optimized Schwarz methods for many subdomains. We investigate here if such a simplified optimization suffices for the magnetotelluric approximation of Maxwell's equation which leads to a complex diffusion problem. We start with a direct analysis for 2 and 3 subdomains, and present asymptotically optimized transmission conditions in each case. We then optimize transmission conditions numerically for 4, 5 and 6 subdomains and observe the same asymptotic behavior of optimized transmission conditions. We finally use the technique of limiting spectra to optimize for a very large number of subdomains in a strip decomposition. Our analysis shows that the asymptotically best choice of transmission conditions is the same in all these situations, only the constants differ slightly. It is therefore enough for such diffusive type approximations of Maxwell's equations, which include the special case of the Laplace and screened Laplace equation, to optimize transmission parameters in the simplified two subdomain decomposition setting to obtain good transmission conditions for optimized Schwarz methods for more general decompositions.
相關內容
Artificial Neuronal Networks are models widely used for many scientific tasks. One of the well-known field of application is the approximation of high-dimensional problems via Deep Learning. In the present paper we investigate the Deep Learning techniques applied to Shape Functionals, and we start from the so--called Torsional Rigidity. Our aim is to feed the Neuronal Network with digital approximations of the planar domains where the Torsion problem (a partial differential equation problem) is defined, and look for a prediction of the value of Torsion. Dealing with images, our choice fell on Convolutional Neural Network (CNN), and we train such a network using reference solutions obtained via Finite Element Method. Then, we tested the network against some well-known properties involving the Torsion as well as an old standing conjecture. In all cases, good approximation properties and accuracies occurred.
Given a positive function $g$ from $[0,1]$ to the reals, the function's missing mass in a sequence of iid samples, defined as the sum of $g(pr(x))$ over the missing letters $x$, is introduced and studied. The missing mass of a function generalizes the classical missing mass, and has several interesting connections to other related estimation problems. Minimax estimation is studied for order-$\alpha$ missing mass ($g(p)=p^{\alpha}$) for both integer and non-integer values of $\alpha$. Exact minimax convergence rates are obtained for the integer case. Concentration is studied for a class of functions and specific results are derived for order-$\alpha$ missing mass and missing Shannon entropy ($g(p)=-p\log p$). Sub-Gaussian tail bounds with near-optimal worst-case variance factors are derived. Two new notions of concentration, named strongly sub-Gamma and filtered sub-Gaussian concentration, are introduced and shown to result in right tail bounds that are better than those obtained from sub-Gaussian concentration.
We study two-user multiple-input single-output (MISO) wireless powered communication networks (WPCNs), where the user devices are equipped with non-linear energy harvesting (EH) circuits. We consider time-division duplex (TDD) transmission, where the users harvest power from the signal received in the downlink phase, and then, utilize this harvested power for information transmission in the uplink phase. In contrast to existing works, we adopt a non-linear model of the harvested power based on a precise analysis of the employed EH circuit. We jointly optimize the beamforming vectors in the downlink and the time allocated for downlink and uplink transmission to minimize the average transmit power in the downlink under per-user data rate constraints in the uplink. We provide conditions for the feasibility of the resource allocation problem and the existence of a trivial solution, respectively. For the case where the resource allocation has a non-trivial solution, we show that it is optimal to employ no more than three beamforming vectors for power transfer in the downlink. To determine these beamforming vectors, we develop an iterative algorithm based on semi-definite relaxation (SDR) and successive convex approximation (SCA). Our simulation results reveal that the proposed resource allocation scheme outperforms two baseline schemes based on linear and sigmoidal EH models, respectively.
We consider a model of energy minimization arising in the study of the mechanical behavior caused by cell contraction within a fibrous biological medium. The macroscopic model is based on the theory of non rank-one convex nonlinear elasticity for phase transitions. We study appropriate numerical approximations based on the discontinuous Galerkin treatment of higher gradients and used succesfully in numerical simulations of experiments. We show that the discrete minimizers converge in the limit to minimizers of the continuous problem. This is achieved by employing the theory of $\Gamma$-convergence of the approximate energy functionals to the continuous model when the discretization parameter tends to zero. The analysis is involved due to the structure of numerical approximations which are defined in spaces with lower regularity than the space where the minimizers of the continuous variational problem are sought. This fact leads to the development of a new approach to $\Gamma$-convergence, appropriate for discontinuous finite element discretizations, which can be applied to quite general energy minimization problems. Furthermore, the adoption of exponential terms penalising the interpenetration of matter requires a new framework based on Orlicz spaces for discontinuous Galerkin methods which is developed in this paper as well.
This paper is concerned with the optimized Schwarz waveform relaxation method and Ventcel transmission conditions for the linear advection-diffusion equation. A mixed formulation is considered in which the flux variable represents both diffusive and advective flux, and Lagrange multipliers are introduced on the interfaces between nonoverlapping subdomains to handle tangential derivatives in the Ventcel conditions. A space-time interface problem is formulated and is solved iteratively. Each iteration involves the solution of time-dependent problems with Ventcel boundary conditions in the subdomains. The subdomain problems are discretized in space by a mixed hybrid finite element method based on the lowest-order Raviart-Thomas space and in time by the backward Euler method. The proposed algorithm is fully implicit and enables different time steps in the subdomains. Numerical results with discontinuous coefficients and various Pecl\'et numbers validate the accuracy of the method with nonconforming time grids and confirm the improved convergence properties of Ventcel conditions over Robin conditions.
This paper is concerned with numerical solution of transport problems in heterogeneous porous media. A semi-discrete continuous-in-time formulation of the linear advection-diffusion equation is obtained by using a mixed hybrid finite element method, in which the flux variable represents both the advective and diffusive flux, and the Lagrange multiplier arising from the hybridization is used for the discretization of the advective term. Based on global-in-time and nonoverlapping domain decomposition, we propose two implicit local time-stepping methods to solve the semi-discrete problem. The first method uses the time-dependent Steklov-Poincar\'e type operator and the second uses the optimized Schwarz waveform relaxation (OSWR) with Robin transmission conditions. For each method, we formulate a space-time interface problem which is solved iteratively. Each iteration involves solving the subdomain problems independently and globally in time; thus, different time steps can be used in the subdomains. The convergence of the fully discrete OSWR algorithm with nonmatching time grids is proved. Numerical results for problems with various Pecl\'et numbers and discontinuous coefficients, including a prototype for the simulation of the underground storage of nuclear waste, are presented to illustrate the performance of the proposed local time-stepping methods.
We propose polynomial-time algorithms to minimise labelled Markov chains whose transition probabilities are not known exactly, have been perturbed, or can only be obtained by sampling. Our algorithms are based on a new notion of an approximate bisimulation quotient, obtained by lumping together states that are exactly bisimilar in a slightly perturbed system. We present experiments that show that our algorithms are able to recover the structure of the bisimulation quotient of the unperturbed system.
As data-driven methods are deployed in real-world settings, the processes that generate the observed data will often react to the decisions of the learner. For example, a data source may have some incentive for the algorithm to provide a particular label (e.g. approve a bank loan), and manipulate their features accordingly. Work in strategic classification and decision-dependent distributions seeks to characterize the closed-loop behavior of deploying learning algorithms by explicitly considering the effect of the classifier on the underlying data distribution. More recently, works in performative prediction seek to classify the closed-loop behavior by considering general properties of the mapping from classifier to data distribution, rather than an explicit form. Building on this notion, we analyze repeated risk minimization as the perturbed trajectories of the gradient flows of performative risk minimization. We consider the case where there may be multiple local minimizers of performative risk, motivated by situations where the initial conditions may have significant impact on the long-term behavior of the system. We provide sufficient conditions to characterize the region of attraction for the various equilibria in this settings. Additionally, we introduce the notion of performative alignment, which provides a geometric condition on the convergence of repeated risk minimization to performative risk minimizers.
In this paper, from a theoretical perspective, we study how powerful graph neural networks (GNNs) can be for learning approximation algorithms for combinatorial problems. To this end, we first establish a new class of GNNs that can solve strictly a wider variety of problems than existing GNNs. Then, we bridge the gap between GNN theory and the theory of distributed local algorithms to theoretically demonstrate that the most powerful GNN can learn approximation algorithms for the minimum dominating set problem and the minimum vertex cover problem with some approximation ratios and that no GNN can perform better than with these ratios. This paper is the first to elucidate approximation ratios of GNNs for combinatorial problems. Furthermore, we prove that adding coloring or weak-coloring to each node feature improves these approximation ratios. This indicates that preprocessing and feature engineering theoretically strengthen model capabilities.
Implicit probabilistic models are models defined naturally in terms of a sampling procedure and often induces a likelihood function that cannot be expressed explicitly. We develop a simple method for estimating parameters in implicit models that does not require knowledge of the form of the likelihood function or any derived quantities, but can be shown to be equivalent to maximizing likelihood under some conditions. Our result holds in the non-asymptotic parametric setting, where both the capacity of the model and the number of data examples are finite. We also demonstrate encouraging experimental results.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191