No abstract uploaded!

In this paper, we present a staggered discontinuous Galerkin immersed boundary method (SDGIBM) for the numerical approximation of fluid-structure interaction. The immersed boundary method is used to model the fluid-structure interaction, while the fluid flow is governed by incompressible Navier-Stokes equations. One advantage of using Galerkin method over the finite difference method with immersed boundary method is that we can avoid approximations of the Dirac Delta function. Another key ingredient of our method is that our solver for incompressible Navier-Stokes equations combines the advantages of discontinuous Galerkin methods and staggered meshes, and results in many good properties, namely local and global conservations and pointwise divergence-free velocity field by a local postprocessing technique. Furthermore, energy stability is improved by a skew-symmetric discretization of the convection term. We will present numerical results to show the performance of the method.

No abstract uploaded!

We establish the existence of infinitely many$polynomial$progressions in the primes; more precisely, given any integer-valued polynomials$P$_{1}, …,$P$_{$k$}∈$Z$[$m$] in one unknown$m$with$P$_{1}(0) = … =$P$_{$k$}(0) = 0, and given any$ε$> 0, we show that there are infinitely many integers$x$and$m$, with $1 \leqslant m \leqslant x^\varepsilon$ , such that$x$+$P$_{1}($m$), …,$x$+$P$_{$k$}($m$) are simultaneously prime. The arguments are based on those in [18], which treated the linear case$P$_{$j$}= ($j$− 1)$m$and$ε$= 1; the main new features are a localization of the shift parameters (and the attendant Gowers norm objects) to both coarse and fine scales, the use of PET induction to linearize the polynomial averaging, and some elementary estimates for the number of points over finite fields in certain algebraic varieties.

In this paper, we aim at developing scalable neural network-type learning systems. Motivated by the idea of constructive neural networks in approximation theory, we focus on constructing rather than training feed-forward neural networks (FNNs) for learning, and propose a novel FNNs learning system called the constructive FNN (CFN). Theoretically, we prove that the proposed method not only overcomes the classical saturation problem for constructive FNN approximation, but also reaches the optimal learning rate when the regression function is smooth, while the state-of-the-art learning rates established for traditional FNNs are only near optimal (up to a logarithmic factor). A series of numerical simulations are provided to show the efficiency and feasibility of CFN.