A5053450580- Topological and Geometric Data Analysis
- Advanced Neuroimaging Techniques and Applications
- Homotopy and Cohomology in Algebraic Topology
- Robotic Locomotion and Control
- Cell Image Analysis Techniques
- Material Dynamics and Properties
- Glass properties and applications
- Biomimetic flight and propulsion mechanisms
- Theoretical and Computational Physics
- Mathematical Dynamics and Fractals
- Computational Drug Discovery Methods
- Protein Structure and Dynamics
- Sports Dynamics and Biomechanics
- Neural Networks and Applications
- Magnetic Properties and Applications
- Bat Biology and Ecology Studies
- Advanced Fluorescence Microscopy Techniques
- Hydrocarbon exploration and reservoir analysis
- Molecular spectroscopy and chirality
- Nuclear Physics and Applications
- X-ray Diffraction in Crystallography
- Machine Learning in Materials Science
- Magnetic properties of thin films
- Leprosy Research and Treatment
- Quantum chaos and dynamical systems
Okayama University
2021-2025
Japan Science and Technology Agency
2013-2023
RIKEN Center for Advanced Intelligence Project
2018-2022
The University of Tokyo
2022
Institute for Molecular Science
2022
Aoyama Gakuin University
2022
The Graduate University for Advanced Studies, SOKENDAI
2022
Tohoku University
2015-2020
Advanced Institute of Materials Science
2018-2020
Gifu University
2017
This paper introduces persistent homology, which is a powerful tool to characterize the shape of data using mathematical concept topology. We explain fundamental idea homology from scratch some examples. also review applications materials researches and software for analysis. HomCloud, one software, especially featured in this paper.
Abstract The broken symmetry in the atomic-scale ordering of glassy versus crystalline solids leads to a daunting challenge provide suitable metrics for describing order within disorder, especially on length scales beyond nearest neighbor that are characterized by rich structural complexity. Here, we address this silica, canonical network-forming glass, using hot cold compression (i) systematically increase after densification and (ii) prepare two glasses with same high-density but...
Abstract Silicate glasses have evolved from basic structural materials to enabling for advanced applications. In this article, we unravel the origin of mixed alkali effect silicate 22.7R 2 O–77.3SiO (R = Na and/or K) by identifying variation in ion location around non-bridging oxygen atoms. To do so, constructed a state-of-the art model, which reproduces both diffraction and NMR data with particular focus on behavior ions. A novel topological analysis using persistent homology found that...
Abstract The purpose of this study is to evaluate the accuracy for classification hepatic tumors by characterization T1-weighted magnetic resonance (MR) images using two radiomics approaches with machine learning models: texture analysis and topological data persistent homology. This assessed non-contrast-enhanced fat-suppressed three-dimensional (3D) 150 tumors. lesions included 50 hepatocellular carcinomas (HCCs), metastatic (MTs), hemangiomas (HHs) found respectively in 37, 23, 33...
We apply a persistent homology analysis to investigate the behavior of nanovoids during crazing process glassy polymers. carry out coarse-grained molecular dynamics simulation uniaxial deformation an amorphous polymer and analyze results with homology. Persistent reveals void coalescence craze formation, suggest that yielding is regarded as percolation created by deformation.
The present paper shows a mathematical formalization of---as well as algorithms and software for computing---volume-optimal cycles. Volume-optimal cycles are useful understanding geometric features appearing in persistence diagram. provide concrete optimal homologous structures, such rings or cavities, on given dataset. key idea is the optimality $(q + 1)$-chain complex $q$th homology generator. This suitable persistent homology. We can solve optimization problem using linear programming....
The structure of glassy, liquid, and amorphous materials is still not well understood, due to the insufficient structural information from diffraction data. In this article, attempts are made understand origin peaks, particularly first sharp peak (FSDP, Q1), principal (PP, Q2), third (Q3), observed in measured patterns disordered whose contains tetrahedral motifs. It confirmed that FSDP (Q1) a signature formation network, because an molecular liquids. found PP (Q2) reflects orientational...
Macroscopic phenomena, such as fracture, corrosion, and degradation of materials, are associated with various reactions which progress heterogeneously. Thus, material properties generally determined not by their averaged characteristics but specific features in heterogeneity (or 'trigger sites') phases, chemical states, etc., where the key that dictate macroscopic initiate propagate. Therefore, identification trigger sites is crucial for controlling properties. However, this a challenging...
High-pressure synthesis of denser glass has been a longstanding interest in condensed-matter physics and materials science because its potentially broad industrial application. Nevertheless, understanding nature under extreme pressures yet to be clarified due experimental theoretical challenges. Here we reveal the formation $\mathrm{OS}{\mathrm{i}}_{4}$ tetraclusters associated with that $\mathrm{Si}{\mathrm{O}}_{7}$ polyhedra $\mathrm{Si}{\mathrm{O}}_{2}$ ultrahigh 200 gigapascal confirmed...
High-accuracy prediction of the physical properties amorphous materials is challenging in condensed-matter physics. A promising method to achieve this machine-learning potentials, which an alternative computationally demanding ab initio calculations. When applying construction descriptors represent atomic configurations crucial. These should be invariant symmetry operations. Handcrafted representations using a smooth overlap positions and graph neural networks (GNN) are examples methods used...
We discuss an algorithmic framework based on efficient graph algorithms and algebraic-topological computational tools. The is aimed at automatic computation of a database global dynamics given m-parameter semidynamical system with discrete time bounded subset the n-dimensional phase space. introduce mathematical background, which upon Conley's topological approach to dynamics, describe for analysis using rectangular grids both in space parameter space, show two sample applications.
Understanding disordered structure is difficult due to insufficient information in experimental data. Here, we overcome this issue by using a combination of diffraction and simulation investigate oxygen packing network topology glassy (g-) liquid (l-) MgO–SiO2 based on comparison with the crystalline topology. We find that atoms Mg2SiO4 larger than MgSiO3, glasses liquids. Moreover, topological analysis suggests similarity between (c)- g-(l-) signature low glass-forming ability (GFA), high...
Abstract The structural origin of the slow dynamics in glass formation remains to be understood owing subtle differences between liquid and states. Even from simulations, where positions all atoms are deterministic, it is difficult extract significant components for formation. In this study, we have extracted local atomic structures a large number metallic models with different cooling rates by utilising computational persistent homology method combined linear machine learning techniques. A...
Abstract Understanding the liquid structure provides information that is crucial to uncovering nature of glass-liquid transition. We apply an aerodynamic levitation technique and high-energy X-rays ( l )-Er 2 O 3 discover its structure. The sample densities are measured by electrostatic at International Space Station. Liquid Er displays a very sharp diffraction peak (principal peak). Applying combined reverse Monte Carlo – molecular dynamics approach, simulations produce Er–O coordination...
Quantifying the correlation between complex structures of amorphous materials and their physical properties has been a longstanding problem in science. In Si, representative covalent solid, presence medium-range order (MRO) intensively discussed. However, specific atomic arrangement corresponding to MRO its relationship with properties, such as thermal conductivity, remains elusive. We solved this by combining topological data analysis, machine learning, molecular dynamics simulations. Using...
Understanding the protein folding process is an outstanding issue in biophysics; recent developments molecular dynamics simulation have provided insights into this phenomenon. However, large freedom of atomic motion hinders understanding process. In study, we applied persistent homology, emerging methods to analyze topological features a dataset, reveal dynamics. We developed new method characterize structure based on homology and simulations chignolin. Using principle component analysis or...
Abstract Topological data analysis is an emerging concept of for characterizing shapes. A state-of-the-art tool in topological persistent homology, which expected to summarize quantified and geometric features. Although homology useful revealing the information, it difficult interpret parameters themselves directly relate physical properties. In this study, we focus on connectivity apertures flow channels detected from analysis. We propose a method estimate permeability fracture networks...
Passive dynamic walking is a useful model for investigating the mechanical functions of body that produce energy-efficient walking. The basin attraction very small and thin, it has fractal-like shape; this explains difficulty in producing stable passive underlying mechanism produces these geometric characteristics was not known. In paper, we consider from viewpoint dynamical systems theory, use simplest to clarify forms We show intrinsic saddle-type hyperbolicity upright equilibrium point...
Persistent homology is a mathematical method to quantify topological features of shapes, such as connectivity. This study applied persistent analyze fracture network patterns in rocks. We show that can detect paths connecting from one boundary the other constituting fractures, which useful for understanding relationships between and flow phenomena. In addition, complex so-called mesh textures serpentine were analyzed by homology. previous studies, different conditions generated...
In this study, we investigate the magnetization process at high frequencies based on energy landscape outputted by machine learning. Employing a combination of Topological Data Analysis (TDA) and learning, analyze how microstructures influence from 1KHz up to 100KHz. Our approach uses magneto-optical Kerr effect (MOKE) microscope for visualizing magnetic domains various frequencies, revealing insights into their behavior structure. Persistent homology, method within TDA, transforms complex...
Passive dynamic walking is a model that walks down shallow slope without any control or input. This has been widely used to investigate how humans walk with low energy consumption and provides design principles for energy-efficient biped robots. However, the basin of attraction very small thin fractal-like complicated shape, which makes producing stable difficult. In our previous study, we simplest investigated based on dynamical systems theory by focusing hybrid dynamics composed continuous...