In this paper, we provide some numerical results for Malliavin sensitivity analysis in finance. We consider payoff functions with polynomial growth, and with and without discontinuous points and compare the results to that of the finite-difference and pathwise methods.

Permutation polynomials over a ring of modulo $2^w$ are well adopted to digital computers and digital signal processors, and so they are in particular expected to be useful for cryptography and pseudo random number generator{s}. For a longer period of the polynomial is demanded in general, we derive a necessary and sufficient condition that polynomials are permutating and their periods are the longest over the ring. We call polynomials which satisfy the condition ``one-stroke polynomials over the ring''.

Estimating the Frobenius norm of a matrix product C=XY without computing C explicitly is required in applications such as the one-sided block Jacobi method. In this paper, we analyze Bečka et al.'s estimator for this problem within a probabilistic framework. Specifically, we consider the set of matrices with the Frobenius norm $\|C\|_F^2$ and introduce some natural probability measure into it. Then, we show that if we choose a matrix randomly from this set and apply the estimator, the expected value of the square of this estimator is exactly $\|C\|_F^2$.

This paper presents a new methodology to compute first-order Greeks for barrier options under the framework of path-dependent payoff functions with European, Lookback, or Asian type and with time-dependent trigger levels. In particular, we develop chain rules for Wiener path integrals between two curves that arise in the computation of first-order Greeks for barrier options. We also illustrate the effectiveness of our method through numerical examples.

The Sakurai-Sugiura method with Rayleigh-Ritz projection (SS-RR method) finds the eigenvalues in a certain domain of the complex plane of large quadratic eigenvalue problems (QEPs). The SS-RR method can suffer from numerical instability when the coefficient matrices of the projected QEP vary widely in norm. To improve the numerical stability of the SS-RR method, we combine it with a numerically stable eigensolver for the small projected QEP. We analyze the backward stability of the proposed method and show, through numerical experiments, that it computes eigenpairs with backward errors that are smaller than those computed by the SS-RR method.

HFE (Patarin, Eurocrypt'96) is one of the most famous multivariate public key cryptosystems. Unfortunately, HFE has a serious trade-off between security and efficiency, which lacks HFE's practicality. Recently, Porras et al. proposed a new encryption scheme ZHFE at PQCrypto 2014. While its construction is similar to HFE, the security seems more than HFE. The present paper proposes a chosen ciphertext attack (CCA) on ZHFE. The CCA reduces the problem of recovering the univariate polynomial for decryption to the min-rank problem on HFE. Thus the CCA security of ZHFE is almost the same as the security of HFE against the min-rank attack.

We propose a mapping operator construction technique for error correction methods applied to a series of linear systems. The mapping operator is constructed based on information obtained in the previous solution step. The additional memory requirement for the construction is limited. Our numerical experiments based on an electromagnetic field analysis confirmed that the correction method with the proposed mapping operator improves the convergence of the iterative solver.

Charles et al. proposed hash functions based on the difficulty of computing isogenies between supersingular elliptic curves. Yoshida and Takashima then improved the 2-isogeny hash function computation by using some specific properties of 2-torsion points. In this paper, we extend the technique to 3-isogenies and give the efficient 3-isogeny hash computation based on a simple representation of the (backtracking) 3-torsion point. Moreover, we implement the 2- and 3-isogeny hash functions using Magma and show our 3-isogeny proposal has a comparable efficiency with the 2-isogeny one.

Hamiltonian Monte Carlo is a Markov chain Monte Carlo method that uses Hamiltonian dynamics to efficiently produce distant samples. It employs geometric numerical integration to simulate Hamiltonian dynamics, which is a key of its high performance. We present a Hamiltonian Monte Carlo method with adaptive step size control to further enhance the efficiency. We propose a new explicit, reversible, and volume-preserving integration method to adaptively set the step sizes, which does not violate the detailed balance condition or require a large increase in computational time.

We carry out a linear stability analysis for a free boundary between two fluids in a Hele-Shaw cell, which is driven by the Darcy's law and an injection or suction at the center. In general, stability of an interface is determined by the well-known Saffman-Taylor instability condition. The linear growth rate of the perturbation depends on two parameters; the rate of the injection/suction and the viscosity contrast of the two fluids. In this paper, we numerically find a parameter region, in which the interface can be stable (resp. unstable) even though it has been considered to be unstable (resp. stable) due to the Saffman-Taylor instability. For the case of the injection, it is suggested that such parameter region vanishes as time increases to infinity. However, the destabilization can be retarded for a sufficiently long time, as one tunes the viscosity and the injection rate of the injecting fluid.

The aim of this paper is to develop the method of fundamental solutions using weighted average condition and dummy points. We accomplish mathematical analysis, a unique existence of an approximate solution and an exponential decay of the approximation error, for a potential problem in disk, and show some numerical experiments, which exemplify our error estimate.

An efficient quadrature formula, known as the Single-Exponential (SE) formula, was proposed by Stenger for the definite integral of an exponentially decaying function over the semi-infinite interval. The formula was derived by combining the trapezoidal formula with a SE transformation. An error bound of the formula was already given. In this study, we investigate another SE formula obtained by replacing the transformation with Muhammad-Mori's SE transformation. Its error bound was determined by theoretical analysis. Numerical comparisons of Stenger's SE formula with that of Muhammad-Mori's are given as well.

This paper proposes advanced credit risk assessment using purchase order information from borrower firms. It first introduces a structural credit risk model based on purchase orders and demonstrates the applicability of the model to practical credit risk monitoring with a case study. The estimated default probabilities reflect trends in purchase order volumes and customers' default risk. The proposed model realizes more frequent credit risk monitoring than typical monitoring based on financial statements. Financial institutions can monitor the actual business conditions of borrower firms using the model.

In continuous time diffusion models, the optimal strategies to utility maximizations can be obtained by solving a certain partial differential equation. In this paper, we give another proof of this fact in an incomplete market without using the well-known fictitious security arguments. Since we avoid using the fictitious security arguments, we can apply our method to the situations when the markets cannot be completed. We provide an example of such cases where the asset price follows a simple jump process with unpredictable jump sizes and see that we can derive the equation which determines the optimal strategy as usual.

We propose a new numerical scheme of evolution for the Einstein equations using the discrete variational derivative method (DVDM). We derive the discrete evolution equation of the constraint using this scheme and show the constraint preserves in the discrete level. In addition, to confirm the numerical stability using this scheme, we perform some numerical simulations by discretized equations with the Crank-Nicolson scheme and with the new scheme, and we find that the new discretized equations have better stability than that of the Crank-Nicolson scheme.

A nonlinear crack problem subject to a non-penetration inequality is considered within the framework of the limiting small strain approach, which does not suffer from the inconsistency of infinite strain at the crack tip. Based on the concept of a generalized solution, sufficient conditions proving the well-posedness of the problem are established and analyzed.

The security of lattice-based cryptography is mainly based on the fact that the shortest vector problem (SVP) is NP-hard. Our interest is to know how large-scale shortest vector problems can be solved by the state-of-the-art software for mixed-integer programs. For this, we provide a formulation for SVP via mixed integer quadratic program and show the numerical performance for TU Darmstadt's benchmark instances with the dimension up to 49.

Truncation errors are considered for approximate differential operators with a class of particle methods. Introducing sufficient conditions for the weight function and a regularity of the family of discrete parameters leads to truncation error estimates of approximate gradient and Laplace operators with a particle method based on the Voronoi decomposition. Moreover, some numerical results agree well with theoretical ones.

Network synthesis problem (NSP) asks to find a minimum-cost network satisfying a given connectivity requirement. Hau, Hirai, and Tsuchimura presented a simple greedy algorithm finding a half-integral optimal solution when the edge-cost is a tree metric. This generalizes the classical result by Gomory and Hu. In this note, we present an integer version of Hau, Hirai, and Tsuchimura's result for integer network synthesis problem (INSP), where a required network must have an integer capacity. We prove that INSP is solvable in polynomial time when the edge-cost is a tree metric and each connectivity requirement is at least 2.

A compactness proof of a nonlinear operator related to stream function-vorticity formulation for the Navier-Stokes equations is presented. The compactness of the operator provides important information for fixed-point formulations, especially for computer-assisted proofs based on Schauder's fixed-point theorem. Our idea for the compactness proof comes from books by Girault & Raviart and Ladyzhenskaia, and our principle would be also applied to convex polygonal regions.

For computing integrals appearing in electronic structure calculations, efficient methods are proposed. The methods consist of two parts: the first part represents the integrals as contour integrals and the second part evaluates the contour integrals using the Clenshaw-Curtis quadrature. The efficiency of the proposed methods is demonstrated through numerical experiments.