• Abstract

Symplecticity is a significant property of stochastic Hamiltonian systems, and the symplectic methods are very attractive. Compared with the classical non-exponential Runge-Kutta methods, the exponential Runge-Kutta methods are more suitable for stiff problems. Therefore, the focus of this paper is on constructing stochastic symplectic exponential Runge-Kutta (SSERK) integrators for semilinear stochastic differential equations (SDEs) driven by multiplicative noise. The first is to establish the symplectic conditions of stochastic exponential Runge-Kutta (SERK) methods. It can be found that when the stiffness matrix is 0, these conditions will degenerate into the symplectic conditions of classical stochastic Runge-Kutta methods. Based on this idea, we construct a class of SSERK integrators with remarkable properties of structure-preservation. In addition, we verify the existence of the quadratic first integral of the stochastic Hamiltonian system and investigate the connection between preserving both the quadratic first integral and the symplectic structure by the SERK methods. Numerical experiments demonstrate a better structure-preserving ability and a higher accuracy of the SSERK integrators in solving the considered semilinear SDEs than the corresponding stochastic symplectic Runge-Kutta integrators. Excitingly, the SSERK integrators perform well when applied to the temporal discretization of stochastic nonlinear Schrödinger equation.

• General Summary

• Abstract

Stomata are important for plants as nodes linking the transpiration process and photosynthesis process. Accurate regulation of stomatal aperture balances water loss and carbon-dioxide uptake. In a recent work, a mathematical model is developed for seed-plant stomata, which exhibits rich dynamical behaviors and can semi-quantitatively explain responses of the stomata to environmental changes. In this work, we numerically study the dynamical behaviors of the stomatal model using phase diagrams with respect to a few important parameters. The phase diagrams are helpful for further validation of the model and experimental measurements of model parameters.

• General Summary

• Abstract

A discontinuous Galerkin (DG) method on Bakhvalov-type ($B$-type) meshes for singularly perturbed Volterra integro-differential equations (SPVIDEs) is proposed. We derive abstract error bounds of the DG method for the SPVIEDs in the $L^2$-norm. It is shown that the approximate solution generated by the DG method on $B$-type meshes has optimal convergence rate $k+1$ in the $L^2$-norm, when using the piecewise polynomial space of degree $k.$ Numerical simulations demonstrate the validity of the theoretical results.

• General Summary

• Abstract

The matrix rank approximation has shown high effectiveness in the matrix rank minimization (MRM) problem, which aims to recover the underlying low-rank structure from the observed matrix by imposing the rank constraint. The nuclear norm, serving as a convex surrogate of matrix rank, is employed in the MRM problem by shrinking singular values of the observed entry. However, this substitution treats each singular value equally, which is virtually $ℓ_1$-norm penalty of the singular value vector. Theoretically, the rank function of the matrix can be considered as $ℓ_0$-norm of its singular values. Consequently, minimizing the nuclear norm frequently results in biased solutions in various applications. In this article, we first propose a novel nonconvex rank approximation, named tight and flexible rank (TFR) approximation, to describe rank function effectively. Specifically, the TFR approximation can more tightly approach the rank function and exhibit greater flexibility in handling diverse singular values, as compared to existing nonconvex rank approximations. Furthermore, we apply TFR approximation to matrix completion and develop a solving algorithm with guaranteed convergence based on the framework of proximal alternating minimization. Extensive experiments reveal that the proposed matrix completion model with TFR approximation outperforms several existing state-of-the-art convex and nonconvex methods.

• General Summary

• Abstract

In this paper, we propose the optimal production transport model, which is an extension of the classical optimal transport model. We observe in economics, the production of the factories can always be adjusted within a certain range, while the classical optimal transport does not take this situation into account. Therefore, differing from the classical optimal transport, one of the marginals is allowed to vary within a certain range in our proposed model. To address this, we introduce a multiple relaxation optimal production transport model and propose the generalized alternating Sinkhorn algorithms, inspired by the Sinkhorn algorithm and the double regularization method. By incorporating multiple relaxation variables and multiple regularization terms, the inequality and capacity constraints in the optimal production transport model are naturally satisfied. Alternating iteration algorithms are derived based on the duality of the regularized model. We also provide a theoretical analysis to guarantee the convergence of our proposed algorithms. Numerical results indicate significant advantages in terms of accuracy and efficiency. Furthermore, we apply the optimal production transport model to the coal production and transport problem. Numerical simulation demonstrates that our proposed model can save the production and transport cost by 13.17%.

• General Summary

• Abstract

Subdiffusion equations with distributed-order fractional derivatives describe important physical phenomena. In this paper, we consider an inverse space-dependent source term problem governed by a distributed order time-fractional diffusion equation using final time data. Based on the series expression of the solution, the inverse source problem can be transformed into a first kind of Fredholm integral equation. The existence, uniqueness and a conditional stability of the considered inverse problem are established. Building upon this foundation, a generalized quasi-boundary value regularization method is proposed to solve the inverse source problem, and then we prove the well-posedness of the regularized problem. Further, we provide the convergence rates of the regularized solution by employing an a priori and an a posteriori regularization parameter choice rule. Numerical examples in one-dimensional and two-dimensional cases are provided to validate the effectiveness of the proposed method.

• General Summary

• Abstract

A molecular beam epitaxial (MBE) model with slope selection consisting of a fourth order Ginzburg-Landau double well potential, is derived from the variation of the free energy. One challenge in constructing the local structure preserving algorithms (SPAs) for the MBE model with slope selection is how to properly discretize the equation in space and time simultaneously in order to preserve the local structure at the discrete level. To resolve this issue, we employ the local energy dissipation property and the energy quadratization techniques. One novelty is that all nonlinear terms are treated semi-explicitly. The other novelty is that we introduce proper intermediate variables to make the space operators act on one single term, which is one crucial step in constructing local SPAs. We then develop two local energy dissipation preserving schemes and show rigorously the local energy dissipation property of the two schemes. Under suitable boundary conditions, such as periodic boundary conditions, the algorithms can preserve not only mass but also global energy dissipation property. Numerical experiments confirm the second-order accuracy and show the excellent performance of the schemes proposed.

• General Summary

• Abstract

This paper concerns the mathematical analyses of the diffusion model in machine learning. The drift term of the backward sampling process is represented as a conditional expectation involving the data distribution and the forward diffusion. The training process aims to find such a drift function by minimizing the mean-squared residue related to the conditional expectation. Using small-time approximations of the Green’s function of the forward diffusion, we show that the analytical mean drift function in DDPM and the score function in SGM asymptotically blow up in the final stages of the sampling process for singular data distributions such as those concentrated on lower-dimensional manifolds, and are therefore difficult to approximate by a network. To overcome this difficulty, we derive a new target function and associated loss, which remains bounded even for singular data distributions. We validate the theoretical findings with several numerical examples.

• General Summary

• Abstract

The main objective of this study is to present a fast and efficient numerical scheme for solving nonlinear integral differential equations with weak singular kernels in three-dimensional domain. First, the temporal derivative and integral term are approximated by the Crank-Nicolson (CN) method and the second-order fractional quadrature rule. After that the spatial discretization is carried out by combining the finite difference method and the alternating direction implicit (ADI) method, and the nonlinear term are approximated using the Taylor expansion. The stability and convergence of the proposed scheme are analyzed, followed by the verification of the theoretical results through numerical experiments.

• General Summary

• Abstract

In this paper, we investigate the blowup mechanism for the 2D isentropic compressible Euler equations with a damping term. Generally, damped Euler equations have a global classical solution for small smooth initial data. However, we identify a class of large initial data for the damped Euler equations, under the condition of azimuthal symmetry and by choosing a suitable self-similar transformation, resulting in the solution to the Cauchy problem blowing up in a finite time. Additionally, we introduce modulation variables to accurately describe the blowup time and location. Furthermore, the blowup profile exhibits a cusp singularity with Hölder $C^{1/3}$ regularity at the blowup point.

• General Summary

• Abstract

In this paper, we introduce energy-stable schemes based on operator splitting methods for Maxwell’s equations in two-dimensional Lorentz dispersive media with transverse electric polarization, namely the sequential splitting scheme (SS-ML) and the Strang-Marchuk splitting scheme (SM-ML). Each splitting scheme involves two substages per time step, where 1D discrete sub-problems are solved using the Crank-Nicolson method for time discretization. Both schemes ensure energy decay and unconditional stability. The convergence analysis reveals that the SS-ML scheme exhibits first-order accuracy in time and second-order accuracy in space based on the energy technique, while the SM-ML scheme achieves second-order accuracy in both time and space. Additionally, numerical dispersion analysis yields two discrete numerical dispersion relation identities for each scheme. Theoretical results are supported by examples and numerical experiments.

• General Summary

• Abstract

The previous studies have shown that the defocusing NLSE has no the modulational instability, and was not found to admit the rogue wave phenomenon so far. In this paper, we address the question of the higher-order rogue wave solutions of the nonlocal $\mathscr{P}\mathscr{T}$-symmetric NLSE in the defocusing regime. Based on Darboux transformation and iterations, we derive an explicit solution for the higher-order rogue waves by adopting a variable separation and Taylor expansion technique. The higher-order rogue wave solutions are expressed in separation-of-variables form. Furthermore, in order to understand these solutions better, patterns of the rogue waves for lowest three order are explored clearly and conveniently. The reported results may be useful for the design of experiments for observation of rogue waves in the defocusing nonlinear physical systems.

• General Summary

• Abstract

A kind of higher-order Boussinesq equation is studied in this work. Based on the Bell polynomials theories, bilinear representations of the equation are derived, and diverse interaction solutions are constructed through Hirota’s bilinear method. These include interaction solutions characterized by hyperbolic cosine and cosine functions, as well as interactions between lump and soliton solutions. In addition, generalized bilinear operators are used in order to construct a new higher-order Boussinesq-like equation, while lump and breather solutions are also developed utilizing Hirota’s bilinear technique. For the various explicit solutions obtained in this work, several of them are considered to selected particular values for the relevant parameters in order to plot different kind of three-dimensional surfaces with associated two-dimensional density profiles to give a comprehensive understanding of the evolution for various solutions.

• General Summary

• Abstract

Multi-view data analysis has gained increasing popularity, in particular multi-view spectral clustering has attracted much attention for its outstanding performance in mining heterogeneity information in multi-view data. However, most spectral clustering methods exhibit the following disadvantages: firstly, learning consensus representation directly from multi-view data that may contain noise renders a distorted description; secondly, the traditional two-step process may fall into a suboptimal solution. To overcome these disadvantages, a novel multi-view spectral clustering method is proposed by unifying the optimization of consensus Laplacian matrix and the learning and discretization of spectral embedding into one step. We consider that the optimal Laplacian matrix is in the neighborhood of view-specific Laplacian matrix, as the view-specific Laplacian matrix only contains partial information from multi-view data, resulting in certain deviation from the optimal Laplacian matrix. The consensus Laplacian matrix was obtained in a dynamic optimization way with the spectral rotation and embedding information simultaneously determined. Extensive experiments have been conducted to demonstrate the effectiveness and superiority of our proposed method.

• General Summary

• Abstract

In this paper, an $hp$ spectral element approximation for distributed optimal control problem governed by an elliptic equation is investigated, whose objective functional does not include the control variable. And the constraint set on control variable is stated with $L^2$-norm. Optimality condition of the continuous and discretized systems are deduced. In order to solve the equivalent systems with high accuracy, $hp$ spectral element method is employed to discretize the constrained optimal control systems. Based on the property of some interpolation operators, a posteriori error estimates are also established by using some properties of some interpolation operators carefully. Finally, a projection gradient algorithm and a numerical example are provided, which confirm our analytical results. Such estimators guarantee the construction of reliable adaptive methods for optimal control problems.

• General Summary