arXiv subjects

Takeo Hoshi

Publications and source records attributed to Takeo Hoshi.

At least 19 recordsLinked to original sources

arXiv2026

ODAT-SE: A modular and parallel framework for simulation-based inverse problems

Open Data Analysis Tool for Science and Engineering (ODAT-SE) is an open-source Python framework for inverse problems that are solved by repeated forward simulations. It succeeds 2DMAT, which was developed for surface-structure analysis with quantum-beam diffraction and has since been applied to other fields. In version 3, the search algorithms were separated from the application-specific solvers, which became independent packages, and the software was renamed; this paper describes the current version 4.0. The framework connects optimization and sampling algorithms to solvers through a common interface for parameter mapping, constraints, and objective evaluation, so that an existing simulation code can be used with any of the algorithms. Relative to the previously published 2DMAT, it provides two-level MPI parallelism, which distributes candidate evaluations over processes and each evaluation over a parallel solver, checkpoint and continuation facilities, random search, global optimizers from SciPy, tensor-train optimization, and a Bayesian workflow that estimates partition-function ratios, model evidence, effective noise levels, and marginal posterior distributions. We validate the Bayesian workflow against closed-form results for a linear regression, benchmark tensor-train optimization on continuous, binary, and mixed-integer problems, and verify the two-level parallelization by the same optimum for all process decompositions. Taking up the positron-diffraction example of the 2DMAT paper, the workflow identifies the correlated structural solutions and an effective noise scale close to the imposed perturbation, while exposing incomplete convergence of symmetry-related mode weights. Solver packages, templates, and a gallery of examples support the use of the framework as a general platform for inverse-problem analysis.

math.NA↗

arXiv2026

Bayesian inference of non-Maxwellian distribution functions from collective Thomson scattering spectra

We investigate Bayesian inference of non-Maxwellian electron distribution functions from collective Thomson scattering (CTS) spectra. We represent distribution functions using multiple candidate models of different complexity and perform inference on synthetic spectra generated from known non-Maxwellian distribution functions. Our analysis demonstrates that a sufficiently flexible model can approximate the overall shape of the ground-truth distribution function. The most plausible model is identified based on model evidence, a statistical measure representing the probability to obtain the observed data given the model. When the candidates include the ground-truth model, the model evidence favors the ground-truth model. Without the ground-truth model as a candidate, the model evidence favors the candidate model with the fewest parameters that adequately approximates the data. The posterior probability density function of the most plausible model reveals the characteristic features and associated uncertainties of the underlying distribution function. The origin of the additional spectral peaks is attributed to a combination of modifications to the dispersion relation and reduced Landau damping. This enables the objective and data-driven identification of distribution functions directly from observed CTS spectra.

physics.plasm-ph↗

arXiv2026

Two-component Bayesian modeling in the data analysis of collective Thomson scattering

We apply Bayesian inference to the analysis of collective Thomson scattering (CTS) spectra, a diagnostic technique for determining local plasma parameters. Using the population annealing Monte Carlo (PAMC) method, we evaluate the posterior probability density functions (PDFs) for six plasma parameters. The analysis is performed separately for the ion acoustic wave (IAW) and electron plasma wave (EPW) features and jointly for both features. The results show that the combined analysis provides constraints on all six parameters, although the degree of constraint depends on the parameter, whereas the separate analyses provide strong constraints only on complementary subsets of the parameters. In particular, the combined analysis provides a clear constraint on the ion charge state, which is not determined by either separate analysis alone. These findings demonstrate the importance of Bayesian modeling that accounts for the two-component nature of plasmas, consisting of ions and electrons.

physics.plasm-ph↗

arXiv2026

Heteroscedastic Neural Surrogate Modeling for Robust and Rapid Bayesian Inference in Fusion Plasma Diagnostics

Bayesian inference via Markov Chain Monte Carlo (MCMC) provides effective parameter estimation, but its real-time application in complex physical systems is hindered by heavy computational bottlenecks and extreme sensitivity to statistical noise. We address this by proposing a neural-network-based probabilistic surrogate framework for rapid and robust MCMC inference. Using fusion plasma Thomson scattering diagnostics as a challenging, noise-dominated testbed, our approach employs a dual-head architecture to simultaneously estimate the expected physical emission spectrum and the channel-wise intrinsic measurement noise variance. By optimizing a Gaussian Negative Log-Likelihood (GNLL) objective, the learned aleatoric uncertainty dynamically buffers the sampler against pathological shot noise. Evaluations demonstrate that this surrogate framework achieves > 1500x acceleration over exact physical forward models, while simultaneously reducing inference error (RMSE) by >20% compared to standard homoscedastic neural baselines, offering a highly promising paradigm for real-time physical analysis.

stat.CO↗

arXiv2025

Conceptual design of Thomson scattering system with high wavelength resolution in magnetically confined plasmas for electron phase-space measurements

We discuss the conceptual design of a spatially-resolved spectroscopy system of Thomson scattering with high wavelength resolution capable of measuring the shape of electron velocity distribution functions in magnetically confined plasmas. We design a spatially-resolved spectrometer with 2560 wavelength channels. The estimated number of scattered photons in a single spectrometer channel is much larger than unity under the experimental setup and plasma parameters of the Compact Helical Device (CHD), indicating sufficient photon statistics for single-shot measurements. Simulations of the scattered spectra show that the signal-to-noise ratio exceeds 5 even under the most unfavorable conditions expected in CHD at full spectral resolution, and further improves with post-processing pixel binning. Bayesian inference applied to the simulated spectra demonstrates that the inferred plasma parameters agree with the input values within the estimated uncertainties. Comparisons between spectra generated from non-Maxwellian electron velocity distribution functions and their Maxwellian fits indicate that deviations from Maxwellian distributions can be identified using the proposed system.

physics.plasm-ph↗

arXiv2025

Monte Carlo simulation method for incoherent Thomson scattering spectra from arbitrary electron distribution functions

We developed a Monte Carlo simulation method to calculate incoherent Thomson scattering spectra in high temperature plasmas. The basic idea is to treat the entire scattering process as the superposition of individual photon-electron interactions. We introduce macro-particles, referred from particle-in-cell simulations, to reduce the computational cost, and obtain scattered spectra within a reasonable computational time. Since the velocity of the interacting electron is randomly sampled from an electron distribution function, the method can be applied to arbitrary electron distribution functions provided an appropriate sampling scheme is available. We present simulation results for relativistic Maxwellian and kappa distribution functions, and compare them with both analytical and numerical spectra for validation. The simulated spectra show good agreement with both analytical and numerical results, demonstrating that the Monte Carlo simulation method can reliably reproduce incoherent Thomson scattering spectra.

physics.plasm-ph↗

arXiv2024

Surface structure of the 3x3-Si phase on Al(111), studied by the multiple usages of positron diffraction and core-level photoemission spectroscopy

The structure of an Al(111)3x3-Si surface was examined by combining data from positron diffraction and core-level photoemission spectroscopy. Analysis of the diffraction rocking curves indicated that the overlayer had a flat honeycomb lattice structure. Simulations of Si core-level spectra calculated via the first-principles indicated that one of the Si atoms in the unit cell was replaced by an Al atom. The surface superstructure was thus a two-dimensional layer of Al-embedded silicene on Al(111).

cond-mat.mtrl-sci↗

arXiv2023

A fast and accurate computation method for reflective diffraction simulations

We present a new computation method for simulating reflection high-energy electron diffraction and the total-reflection high-energy positron diffraction experiments. The two experiments are used commonly for the structural analysis of material surface. The present paper improves the conventional numerical method, the multi-slice method, for faster computation, since the present method avoids the matrix-eigenvalue solver for the computation of matrix exponentials and can adopt higher-order ordinary differential equation solvers. Moreover, we propose a high-performance implementation based on multi-thread parallelization and cache-reusable subroutines. In our tests, this new method performs up to 2,000 times faster than the conventional method.

math.NA↗

arXiv2022

Data-analysis software framework 2DMAT and its application to experimental measurements for two-dimensional material structures

An open-source data-analysis framework 2DMAT has been developed for experimental measurements of two-dimensional material structures. 2DMAT offers five analysis methods: (i) Nelder-Mead optimization, (ii) grid search, (iii) Bayesian optimization, (iv) replica exchange Monte Carlo method, and (v) population-annealing Monte Carlo method. Methods (ii) through (v) are implemented by parallel computation,which is efficient not only for personal computers but also for supercomputers.The current version of 2DMAT is applicable to total-reflection high-energy positron diffraction (TRHEPD), surface X-ray diffraction (SXRD), and low-energy electron diffraction (LEED) experiments by installing corresponding forward problem solvers that generate diffraction intensity data from a given dataset of the atomic positions. The analysis methods are general and can be applied also to other experiments and problems.

cond-mat.mtrl-sci↗

arXiv2022

sim-trhepd-rheed -- Open-source simulator of total-reflection high-energy positron diffraction (TRHEPD) and reflection high-energy electron diffraction (RHEED)

The present paper reports sim-trhepd-rheed (STR), an open-source simulator of total-reflection high-energy positron diffraction (TRHEPD) and reflection high-energy electron diffraction (RHEED) experiments which are used for atom-scale surface structure determination of a material. The STR simulator is used for the analysis of experimental diffraction data by simulating the rocking curve from a given trial surface structure by solving the partial differential equation of the dynamical quantum diffraction theory for positron or electron wavefunctions. Using the obtained surface structure, electronic structure, and other physical quantities can be evaluated through first-principles calculations. For this purpose, a utility software was also developed in order to realize a first principles calculation with the Quantum ESPRESSO suite.

cond-mat.mtrl-sci↗

arXiv2021

Performance prediction of massively parallel computation by Bayesian inference

A performance prediction method for massively parallel computation is proposed. The method is based on performance modeling and Bayesian inference to predict elapsed time T as a function of the number of used nodes P (T=T(P)). The focus is on extrapolation for larger values of P from the perspective of application researchers. The proposed method has several improvements over the method developed in a previous paper, and application to real-symmetric generalized eigenvalue problem shows promising prediction results. The method is generalizable and applicable to many other computations.

math.NA↗

arXiv2021

Data-driven sensitivity analysis in a total-reflection high-energy positron diffraction (TRHEPD) experiment of the Si$_4$O$_5$N$_3$ / 6H-SiC (0001)-($\sqrt{3} \times \sqrt{3}$) R30$^\circ$

The present article proposes a data analysis method for experimentally-derived measurements, which consists of an auto-optimization procedure and a sensitivity analysis. The method was applied to the results of a total-reflection high-energy positron diffraction (TRHEPD) experiment, a novel technique of determining surface structures or the position of the atoms near the material surface. This method solves numerically the partial differential equation in the fully-dynamical quantum diffraction theory with many trial surface structures. In the sensitivity analysis, we focused on the experimental uncertainties and the variation over individual fitting parameters, which was analyzed by solving the eigenvalue problem of the variance-covariance matrix. A modern massively parallel supercomputer was used to complete the analysis within a moderate computational time. The sensitivity analysis provides a basis for the choice of variables in the data analysis for practical reliability. The effectiveness of the present analysis method was demonstrated in the structure determination of a Si$_4$O$_5$N$_3$ / 6H-SiC(0001)-($\sqrt{3} \times \sqrt{3}$) R30$^\circ$ surface. Furthermore, this analysis method is applicable to many experiments other than TRHEPD.

cond-mat.mtrl-sci↗

arXiv2020

A two-stage data-analysis method for total-reflection high-energy positron diffraction (TRHEPD)

Total-reflection high-energy positron diffraction (TRHEPD) is a novel experimental method for the determination of surface structure, which has been extensively developed at the Slow Positron Facility, Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK). In this paper, a two-stage data-analysis method is proposed. The data analysis is based on an inverse problem in which the atomic positions of a surface structure are determined from the experimental diffraction data (rocking curves). The relevant forward problem is solved by the numerical solution of the partial differential equation for quantum scattering of the positron. In the present two-stage method, the first stage is a grid-based global search and the second stage is a local search for the unique candidate for the atomic arrangement. The numerical problem is solved on a supercomputer

cond-mat.mtrl-sci↗

arXiv2020

K$ω$ -- Open-source library for the shifted Krylov subspace method of the form $(zI-H)x=b$

We develop K$ω$, an open-source linear algebra library for the shifted Krylov subspace methods. The methods solve a set of shifted linear equations $(z_k I-H)x^{(k)}=b\, (k=0,1,2,...)$ for a given matrix $H$ and a vector $b$, simultaneously. The leading order of the operational cost is the same as that for a single equation. The shift invariance of the Krylov subspace is the mathematical foundation of the shifted Krylov subspace methods. Applications in materials science are presented to demonstrate the advantages of the algorithm over the standard Krylov subspace methods such as the Lanczos method. We introduce benchmark calculations of (i) an excited (optical) spectrum and (ii) intermediate eigenvalues by the contour integral on the complex plane. In combination with the quantum lattice solver $\mathcal{H} Φ$, K$ω$ can realize parallel computation of excitation spectra and intermediate eigenvalues for various quantum lattice models.

math.NA↗

arXiv2020

White Paper from Workshop on Large-scale Parallel Numerical Computing Technology (LSPANC 2020): HPC and Computer Arithmetic toward Minimal-Precision Computing

In numerical computations, precision of floating-point computations is a key factor to determine the performance (speed and energy-efficiency) as well as the reliability (accuracy and reproducibility). However, precision generally plays a contrary role for both. Therefore, the ultimate concept for maximizing both at the same time is the minimal-precision computing through precision-tuning, which adjusts the optimal precision for each operation and data. Several studies have been already conducted for it so far (e.g. Precimoniuos and Verrou), but the scope of those studies is limited to the precision-tuning alone. Hence, we aim to propose a broader concept of the minimal-precision computing system with precision-tuning, involving both hardware and software stack. In 2019, we have started the Minimal-Precision Computing project to propose a more broad concept of the minimal-precision computing system with precision-tuning, involving both hardware and software stack. Specifically, our system combines (1) a precision-tuning method based on Discrete Stochastic Arithmetic (DSA), (2) arbitrary-precision arithmetic libraries, (3) fast and accurate numerical libraries, and (4) Field-Programmable Gate Array (FPGA) with High-Level Synthesis (HLS). In this white paper, we aim to provide an overview of various technologies related to minimal- and mixed-precision, to outline the future direction of the project, as well as to discuss current challenges together with our project members and guest speakers at the LSPANC 2020 workshop; https://www.r-ccs.riken.jp/labs/lpnctrt/lspanc2020jan/.

cs.DC↗

arXiv2020

A Parallelizable Energy-Preserving Integrator MB4 and Its Application to Quantum-Mechanical Wavepacket Dynamics

In simulating physical systems, conservation of the total energy is often essential, especially when energy conversion between different forms of energy occurs frequently. Recently, a new fourth order energy-preserving integrator named MB4 was proposed based on the so-called continuous stage Runge--Kutta methods (Y.~Miyatake and J.~C.~Butcher, SIAM J.~Numer.~Anal., 54(3), 1993-2013). A salient feature of this method is that it is parallelizable, which makes its computational time for one time step comparable to that of second order methods. In this paper, we illustrate how to apply the MB4 method to a concrete ordinary differential equation using the nonlinear Schrödinger-type equation on a two-dimensional grid as an example. This system is a prototypical model of two-dimensional disordered organic material and is difficult to solve with standard methods like the classical Runge--Kutta methods due to the nonlinearity and the $δ$-function like potential coming from defects. Numerical tests show that the method can solve the equation stably and preserves the total energy to 16-digit accuracy throughout the simulation. It is also shown that parallelization of the method yields up to 2.8 times speedup using 3 computational nodes.

math.NA↗

arXiv2020

An a posteriori verification method for generalized real-symmetric eigenvalue problems in large-scale electronic state calculations

An a posteriori verification method is proposed for the generalized real-symmetric eigenvalue problem and is applied to densely clustered eigenvalue problems in large-scale electronic state calculations. The proposed method is realized by a two-stage process in which the approximate solution is computed by existing numerical libraries and is then verified in a moderate computational time. The procedure returns intervals containing one exact eigenvalue in each interval. Test calculations were carried out for organic device materials, and the verification method confirms that all exact eigenvalues are well separated in the obtained intervals. This verification method will be integrated into EigenKernel (https://github.com/eigenkernel/), which is middleware for various parallel solvers for the generalized eigenvalue problem. Such an a posteriori verification method will be important in future computational science.

physics.comp-ph↗

arXiv2019

Development of data-analysis software for total-reflection high-energy positron diffraction (TRHEPD)

The present paper reports on the recent activity of the data analysis software development for total-reflection high-energy positron diffraction (TRHEPD), a novel experimental technique for surface structure determination. Experiments using TRHEPD are being conducted intensively at the Slow Positron Facility, Institute of Materials Structure Science, High Energy Accelerator Research Organization, revealing surface structure of interest. The data analysis software provides a solution to the inverse problem in which the atomic positions of a surface structure are determined from the experimental diffraction data (rocking curve). The forward problem is solved by the numerical solution of the partial differential equation in the quantum scattering problem. A technical demonstration with a test problem was carried out to confirm the software functioned as expected. Since the analysis method has a general mathematical foundation, it is also applicable to other experiments, such as X-ray or electron diffraction experiments.

cond-mat.mtrl-sci↗