SEARCH
検索詳細
谷口 隆晴大学院理学研究科 数学専攻教授
研究者基本情報
■ 学位■ 研究ニュース
- 2021年12月09日, さまざまなデータから隠れた物理法則を見つける人工知能 ~物理シミュレーションの新たな応用の可能性に期待~
- 2020年12月08日, 物理法則に忠実なシミュレーションを行う人工知能 ~デジタル解析学でエネルギー保存・減衰性を再現~
- 2019年10月04日, 谷口隆晴准教授がJST(科学技術振興機構)戦略的創造研究推進事業(CREST)に、杉本泰助教、富樫英助教が戦略的創造研究推進事業(さきがけ)に採択されました
■ 研究分野
■ 委員歴
- 2021年04月 - 現在, MDPI Mathematics, Topic Editor
- 2015年05月 - 2024年03月, 日本学術会議, 計算音響学小委員会 委員
- 2019年10月 - 2021年09月, 日本数学会応用数学分科会委員会委員
- 2018年04月 - 2021年03月, 日本応用数理学会, JSIAM Letters 幹事編集委員長
- 2015年09月 - 2018年03月, 日本応用数理学会, JSIAM Letters 副編集委員長
- 2015年04月 - 2017年03月, 日本応用数理学会, 若手の会 幹事
- 28th International Conference on Artificial Neural Networks, Programme Committee
研究活動情報
■ 受賞- 2024年10月 神戸大学, 学長表彰(財務貢献者)
- 2023年10月 神戸大学, 学長表彰(財務貢献者)
- 2023年09月 日本応用数理学会, JSIAM Letters 論文賞, Causal inference for empirical dynamical systems based on persistent homology
- 2021年08月 日本応用数理学会, 日本応用数理学会論文賞 理論部門, 波動方程式と弾性方程式からなる連成系のシンプレクティック性について
- 2017年09月 日本応用数理学会, 日本応用数理学会論文賞(理論部門), ハミルトン方程式に対する離散勾配法のRiemann構造不変性学会誌・学術雑誌による顕彰
- 2016年06月 日本応用数理学会, 日本応用数理学会研究部会連合発表会優秀講演賞, 第12回日本応用数理学会研究部会連合発表会における講演「自動離散微分とその応用」国内学会・会議・シンポジウム等の賞
- 2014年09月 日本応用数理学会, 日本応用数理学会論文賞(理論部門), コンパクト差分に基づく離散変分導関数法
- 2012年08月 日本応用数理学会, 日本応用数理学会若手優秀講演賞, ホロノミック系に対するラグランジュ力学的離散勾配法
- 2011年07月 SciCADE 2011 (the International Conference on Scientific Computation And Differential Equations 2011), SciCADE 2011 New Talent Award, A Lagrangian Approach to Deriving Energy-Preserving Numerical Schemes for the Euler-Lagrange Partial Differential Equations and Its Applications
- 2025年05月, Proc. of the 28th International Conference on Artificial Intelligence and Statistics (AISTATS2025), 英語Energy-Consistent Neural Operators for Hamiltonian and Dissipative Partial Differential Equations[査読有り]研究論文(国際会議プロシーディングス)
- 2025年04月, Proc. of the Thirteenth International Conference on Learning Representations (ICLR2025), 英語Poisson-Dirac Neural Networks for Modeling Coupled Dynamical Systems across Domains[査読有り]研究論文(国際会議プロシーディングス)
- 2025年02月, Proc. of the 39th Annual AAAI Conference on Artificial Intelligence(AAAI2025), 英語Number Theoretic Accelerated Learning of Physics-Informed Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- 2025年, IEEE Transactions on Neural Networks and Learning Systems研究論文(学術雑誌)
- null, 2024年12月, NeurIPS 2024 Workshop on Machine Learning and the Physical Sciences, 英語Port-Hamiltonian Neural Networks for Learning Coupled Systems and Their Interactions[査読有り]研究論文(国際会議プロシーディングス)
- null, 2024年12月, Proc. of 2024 International Symposium on Nonlinear Theory and Its Applications (NOLTA2024), 英語Learning Difference and Summation Operators for Discretization of Nonlocal Hamiltonian Partial Differential Equations Using Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- null, 2024年12月, Proc. of 2024 International Symposium on Nonlinear Theory and Its Applications (NOLTA2024), 英語Application of the Kernel Method to Learning Symplectic Forms[査読有り]研究論文(国際会議プロシーディングス)
- null, 2024年12月, Proc. of 2024 International Symposium on Nonlinear Theory and Its Applications (NOLTA2024), 英語A New Approach to Designing Robust Hamiltonian Neural Networks by Regularisation[査読有り]研究論文(国際会議プロシーディングス)
- null, 2024年12月, Proc. of 2024 International Symposium on Nonlinear Theory and Its Applications (NOLTA2024), 英語Hyperbolic-PDE-Based Neural Network Architecture[査読有り]研究論文(国際会議プロシーディングス)
- Elsevier BV, 2024年10月, Physica D: Nonlinear Phenomena, 134382 - 134382, 英語[査読有り]研究論文(学術雑誌)
- null, 2024年07月, IEICE Transactions on Information and Systems, E107-D, 英語Loss Function for Deep Learning to Model Dynamical Systems[査読有り]研究論文(学術雑誌)
- null, 2024年06月, Proc. of CAI2024 Workshop on Scientific Machine Learning and Its Industrial Applications (SMLIA2024), 英語Improved input points estimate for identifying nonlinear dynamic systems in DeepONet[査読有り]研究論文(国際会議プロシーディングス)
- null, 2024年06月, Proc. of CAI2024 Workshop on Scientific Machine Learning and Its Industrial Applications (SMLIA2024), 英語Learning Coupled Systems and their Connectivity Using Port-Hamiltonian Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- 2024年04月, Mathematics, 英語[査読有り]研究論文(学術雑誌)
- Estimating governing equations from observed time-series data is crucial for understanding dynamical systems. From the perspective of system comprehension, the demand for accurate estimation and interpretable results has been particularly emphasized. Herein, we propose a novel data-driven method for estimating the governing equations of dynamical systems based on machine learning with high accuracy and interpretability. The proposed method enhances the estimation accuracy for dynamical systems using sparse modeling by incorporating physical constraints derived from Hamiltonian mechanics. Unlike conventional approaches used for estimating governing equations for dynamical systems, we employ a sparse representation of Hamiltonian, allowing for the estimation. Using noisy observational data, the proposed method demonstrates a capability to achieve accurate parameter estimation and extraction of essential nonlinear terms. In addition, it is shown that estimations based on energy conservation principles exhibit superior accuracy in long-term predictions. These results collectively indicate that the proposed method accurately estimates dynamical systems while maintaining interpretability.MDPI AG, 2024年03月, Mathematics, 12(7) (7), 974 - 974, 英語[査読有り]研究論文(学術雑誌)
- 2023年12月, NeurIPS2023 Workshop: Machine Learning with New Compute Paradigms, 英語, 国際誌, 国際共著していないAlgebraic Design of Physical Computing System for Time-Series Generation[査読有り]研究論文(国際会議プロシーディングス)
- 2023年09月, IEICE Proceedings Series, 76, 419 - 421, 英語, 国際誌, 国際共著していないApplication of the Neural Operator for Physical Simulations of GENERIC Systems[査読有り]研究論文(国際会議プロシーディングス)
- 2023年09月, IEICE Proceedings Series, 76, 370 - 373, 英語, 国際誌, 国際共著していないSuper Resolution of Numerical Solutions of Nonlinear Elliptic Equations by DeepONet[査読有り]研究論文(国際会議プロシーディングス)
- 2023年09月, IEICE Proceedings Series, 76, 259 - 262, 英語, 国際誌, 国際共著していないGeneralization Error Analysis of Discrete Hamiltonian Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- 2023年07月, ICML Workshop on New Frontiers in Learning, Control, and Dynamical Systems, 英語, 国際誌, 国際共著していないVariational Principle and Variational Integrators for Neural Symplectic Forms[査読有り]研究論文(国際会議プロシーディングス)
- 2023年07月, ICML Workshop on New Frontiers in Learning, Control, and Dynamical Systems, 英語, 国際誌, 国際共著していないEquivalence Class Learning for GENERIC Systems[査読有り]研究論文(国際会議プロシーディングス)
- 2023年07月, 1st Workshop on the Synergy of Scientific and Machine Learning Modeling at ICML2023, 英語, 国際誌, 国際共著していないGood Lattice Accelerates Physics-Informed Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- 2023年05月, Proc. of The Eleventh International Conference on Learning Representations (ICLR2023), 11, 英語, 国際誌, 国際共著していないFINDE: Neural Differential Equations for Finding and Preserving Invariant Quantities[査読有り]研究論文(国際会議プロシーディングス)
- Institute of Electrical and Electronics Engineers (IEEE), 2023年, IEEE Transactions on Neural Networks and Learning Systems, 1 - 13研究論文(学術雑誌)
- 2022年10月, 応用物理, 91(10) (10), 629 - 633, 日本語幾何学的深層科学技術計算 -深層学習による物理モデリング・シミュレーション-[招待有り]研究論文(学術雑誌)
- The Japan Society for Industrial and Applied Mathematics, 2022年03月, JSIAM Letters, 14, 37 - 40, 英語[査読有り]研究論文(学術雑誌)
- 2022年02月, Thirty-Sixth AAAI Conference on Artificial Intelligence, 英語KAM Theory Meets Statistical Learning Theory: Hamiltonian Neural Networks with Non-Zero Training Loss[査読有り]研究論文(国際会議プロシーディングス)
- 2022年, Proceedings of the 2022 International Symposium on Nonlinear Theory and its Applications (NOLTA2022)Learning GENERIC Systems Using Neural Symplectic Forms[査読有り]
- 2022年, Proceedings of the 2022 International Symposium on Nonlinear Theory and its Applications (NOLTA2022), 英語Variational Integrator for Hamiltonian Neural Networks[査読有り]研究論文(国際会議プロシーディングス)
- 2022年, Journal of Signal Processing, 日本語Secure Communication Systems Based on Synchronization of Chaotic Vibration of Wave Equations[査読有り]研究論文(学術雑誌)
- 2022年, ICLR2022 Workshop on AI for Earth and Space Science (ai4earth), 英語Imbalance-Aware Learning for Deep Physics Modeling[査読有り]研究論文(国際会議プロシーディングス)
- 2021年12月, Advances in Neural Information Processing Systems (NeurIPS), 34, 英語Neural Symplectic Form: Learning Hamiltonian Equations on General Coordinate Systems[査読有り]研究論文(国際会議プロシーディングス)
- 2021年12月, Advances in Neural Information Processing Systems (NeurIPS), 34, 英語Symplectic Adjoint Method for Exact Gradient of Neural ODE with Minimal Memory[査読有り]研究論文(国際会議プロシーディングス)
- In a secret communication system using chaotic synchronization, the communication information is embedded in a signal that behaves as chaos and is sent to the receiver to retrieve the information. In a previous study, a chaotic synchronous system was developed by integrating the wave equation with the van der Pol boundary condition, of which the number of the parameters are only three, which is not enough for security. In this study, we replace the nonlinear boundary condition with an artificial neural network, thereby making the transmitted information difficult to leak. The neural network is divided into two parts; the first half is used as the left boundary condition of the wave equation and the second half is used as that on the right boundary, thus replacing the original nonlinear boundary condition. We also show the results for both monochrome and color images and evaluate the security performance. In particular, it is shown that the encrypted images are almost identical regardless of the input images. The learning performance of the neural network is also investigated. The calculated Lyapunov exponent shows that the learned neural network causes some chaotic vibration effect. The information in the original image is completely invisible when viewed through the image obtained after being concealed by the proposed system. Some security tests are also performed. The proposed method is designed in such a way that the transmitted images are encrypted into almost identical images of waves, thereby preventing the retrieval of information from the original image. The numerical results show that the encrypted images are certainly almost identical, which supports the security of the proposed method. Some security tests are also performed. The proposed method is designed in such a way that the transmitted images are encrypted into almost identical images of waves, thereby preventing the retrieval of information from the original image. The numerical results show that the encrypted images are certainly almost identical, which supports the security of the proposed method.MDPI AG, 2021年07月, Entropy, 23(7) (7), 904 - 904, 英語[査読有り]研究論文(学術雑誌)
- Institute of Electronics, Information and Communications Engineers (IEICE), 2021年07月, Nonlinear Theory and Its Applications, IEICE, 12(3) (3), 295 - 308, 英語[査読有り][招待有り]研究論文(学術雑誌)
- 2021年05月, ICLR2021 Workshop on Deep Learning for Simulation (SimDL),, 英語Deep Discrete- Time Lagrangian Mechanics[査読有り]研究論文(国際会議プロシーディングス)
- 2021年02月, 計測自動制御学会論文集, 57(2) (2), 78 - 85, 日本語分布系のカオス同期化を用いた秘匿通信システム[査読有り]研究論文(学術雑誌)
- IEEE, 2021年01月, Proceedings of 2021 IEEE 18th Annual Consumer Communications & Networking Conference (CCNC), 1 - 4, 英語[査読有り]研究論文(国際会議プロシーディングス)
- 2020年12月, 日本応用数理学会論文誌, 30(4) (4), 269 - 289, 日本語波動方程式と弾性方程式からなる連成系のシンプレクティック性について[査読有り]研究論文(学術雑誌)
- 2020年12月, Advances in Neural Information Processing Systems (NeurIPS), 33, 13100 - 13111, 英語Deep Energy-Based Modeling of Discrete-Time Physics[査読有り]研究論文(国際会議プロシーディングス)
- 2020年11月, Proceedings of the 2020 International Symposium on Nonlinear Theory and its Applications (NOLTA2020), 204 - 207, 英語Parameter estimation for dynamical systems via structural realization[査読有り]研究論文(国際会議プロシーディングス)
- Recently, soft robots that consist of soft and deformable materials have received much attention for their adaptability to uncertain environments. Although these robots are difficult to control with a conventional control theory owing to their complex body dynamics, research from different perspectives attempts to actively exploit these body dynamics as an asset rather than a drawback. This approach is called morphological computation, in which the soft materials are used for computation that includes a new kind of control strategy. In this article, we propose a novel approach to analyze the computational properties of soft materials based on an algebraic method, called the input–output equation used in systems analysis, particularly in systems biology. We mainly focus on the two scenarios relevant to soft robotics, that is, analysis of the computational capabilities of soft materials and design of the input force to soft devices to generate the target behaviors. The input–output equation directly describes the relationship between inputs and outputs of a system, and hence by using this equation, important properties, such as the echo state property that guarantees reproducible responses against the same input stream, can be investigated for soft structures. Several application scenarios of our proposed method are demonstrated using typical soft robotic settings in detail, including linear/nonlinear models and hydrogels driven by chemical reactions.SAGE Publications, 2020年03月, The International Journal of Robotics Research, 40(1) (1), 027836492091229 - 027836492091229, 英語[査読有り]研究論文(学術雑誌)
- {MDPI} {AG}, 2020年02月, Mathematics, 8(2) (2), 249 - 249, 英語[査読有り][招待有り]研究論文(学術雑誌)
- 2019年12月, Proceedings of the 2019 International Symposium on Nonlinear Theory and its Applications (NOLTA2019), 187 - 190, 英語Differential Algebraic Method for Direct Evaluation of Computational Capabilities of Physical Reservoirs[査読有り]研究論文(国際会議プロシーディングス)
- 2019年01月, Japan Journal of Industrial and Applied Mathematics, 36(1) (1), 3 - -24, 英語[査読有り]研究論文(学術雑誌)
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2018年, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 11141 LNCS, 781 - 794, 英語[査読有り]研究論文(学術雑誌)
- 2018年, 日本応用数理学会論文誌, 28, 162 - 204, 日本語Husbyらの実験データに対するアレルギー発症メカニズムの解析に向けた抗原・抗体の体内動態モデルの構築[査読有り]研究論文(学術雑誌)
- 2018年01月, Japan Journal of Industrial and Applied Mathematics, 35(2) (2), 英語[査読有り]研究論文(学術雑誌)
- 2017年10月, GERIATRICS & GERONTOLOGY INTERNATIONAL, 17(10) (10), 1752 - 1758, 英語[査読有り]研究論文(学術雑誌)
- 日本応用数理学会 ; 1991-, 2017年03月, 応用数理, 27(1) (1), 13 - 20, 日本語[査読有り][招待有り]研究論文(学術雑誌)
- 2017年03月, MI Lecture Notes of IMI, 74, 31 - 33, 英語Geometric-mechanics-inspired model of stochastic dynamical systems研究論文(国際会議プロシーディングス)
- 2017年03月, MI Lecture Notes of IMI, 74, 63 - 68, 英語Energy-preserving Discrete Gradient Schemes for the Hamilton Equation Based on the Variational Principle研究論文(国際会議プロシーディングス)
- 神戸大学経済経営学会, 2016年11月, 國民經濟雜誌, 214(5) (5), 39 - 50, 日本語研究論文(学術雑誌)
- 一般社団法人 日本応用数理学会, 2016年09月, JSIAM Letters, 8, 53 - 56, 英語
In this contribution, we propose a new framework to derive energy-preserving numerical schemes based on the variational principle for Hamiltonian mechanics. We focus on Noether's theorem, which shows that the symmetry with respect to time translation gives the energy conservation law. By reproducing the calculation of the proof of Noether's theorem after discretization using the summation by parts and the discrete gradient, we obtain the scheme and the corresponding discrete energy at the same time. The significant property of efficiency is that the appropriate choice of the discrete gradient makes our schemes explicit if the Hamiltonian is separable.
[査読有り]研究論文(学術雑誌) - 2015年12月, 2015年度応用数学合同研究集会予稿集, 394 - 401, 日本語地域コミュニティの構造変化に対する検定理論研究論文(研究会,シンポジウム資料等)
- 2015年, PROCEEDINGS OF THE INTERNATIONAL CONFERENCE OF NUMERICAL ANALYSIS AND APPLIED MATHEMATICS 2014 (ICNAAM-2014), 1648, 英語[査読有り]研究論文(国際会議プロシーディングス)
- 2015年, PROCEEDINGS OF THE INTERNATIONAL CONFERENCE OF NUMERICAL ANALYSIS AND APPLIED MATHEMATICS 2014 (ICNAAM-2014), 1648, 英語[査読有り]研究論文(国際会議プロシーディングス)
- We consider application of the discrete gradient method for the Webster equation, which models sound waves in tubes. Typically Hamilton equations are described by the use of gradients of the Hamiltonian and it is indispensable to introduce an inner product to define a gradient. We first apply the discrete gradient method to design an energy-preserving method by using a weighted inner product. Comparing with another scheme that is derived by a standard inner product, we show that the discrete gradient method has a geometric invariance, which implies that the method reflects the symplectic geometric aspect of mechanics.The Japan Society for Industrial and Applied Mathematics, 2015年01月, JSIAM Letters, 7, 17 - 20, 英語[査読有り]研究論文(学術雑誌)
- 2013年09月, ESAIM-MATHEMATICAL MODELLING AND NUMERICAL ANALYSIS-MODELISATION MATHEMATIQUE ET ANALYSE NUMERIQUE, 47(5) (5), 1493 - 1513, 英語[査読有り]研究論文(学術雑誌)
- ソリトン方程式のように保存量を持つ偏微分方程式に対しては,それらを保つ「構造保存数値解法」が適している.一方数値流体の分野では,波動現象の記述に適した「コンパクト差分法」がよく用いられる.本論文では,構造保存数値解法の一種である「離散変分導関数法」において両手法を組み合わせられること,すなわちコンパクト差分に基づく離散変分導関数法を構成できることを述べ,数値例によりその有効性を示す.一般社団法人 日本応用数理学会, 2013年06月, 日本応用数理学会論文誌, 23(2) (2), 203 - 232, 日本語[査読有り]研究論文(学術雑誌)
- 本論文ではEuler-Lagrange偏微分方程式に対し,局所的エネルギー保存則を保つ有限差分スキーム導出法を提案する.この保存則はLagrangianの局所的時間対称性から導出されるが,本論文では,この対称性からEuler-Lagrange方程式自体も導出できることに着目し,これと離散勾配法を組み合わせる.応用として,線形波動方程式に対する無反射境界条件の離散化法についても論じる.一般社団法人 日本応用数理学会, 2012年09月, 日本応用数理学会論文誌, 22(3) (3), 143 - 169, 日本語[査読有り]研究論文(学術雑誌)
- 2012年05月, JOURNAL OF COMPUTATIONAL PHYSICS, 231(10) (10), 3963 - 3986, 英語[査読有り]研究論文(学術雑誌)
- 2012年05月, JOURNAL OF COMPUTATIONAL PHYSICS, 231(14) (14), 4542 - 4559, 英語[査読有り]研究論文(学術雑誌)
- We propose a new structure-preserving integrator for the Korteweg-de Vries (KdV) equation. In this integrator, two independent structure-preserving techniques are newly combined; the "discrete variational derivative method" for constructing invariants-preserving integrator, and the "compact finite difference method" which is widely used in the area of numerical fluid dynamics for resolving wave propagation phenomena. Numerical experiments show that the new integrator is in fact advantageous than the existing integrators.The Japan Society for Industrial and Applied Mathematics, 2012年03月, JSIAM Letters, vol. 4, 5-8., 5 - 8, 英語[査読有り]研究論文(学術雑誌)
- 2011年07月, JOURNAL OF PHYSICS A-MATHEMATICAL AND THEORETICAL, 44(30) (30), 英語[査読有り]研究論文(学術雑誌)
- We consider structure-preserving integration of the Ostrovsky equation, which for example models gravity waves under the influence of Coriolis force. We find a multi-symplectic formulation, and derive a finite difference discretization based on the formulation and by means of the Preissman box scheme. We also present a numerical example, which shows the effectiveness of this scheme.The Japan Society for Industrial and Applied Mathematics, 2011年06月, JSIAM Letters, vol. 3, 41-44., 41 - 44, 英語[査読有り]研究論文(学術雑誌)
- 2010年12月, JAPAN JOURNAL OF INDUSTRIAL AND APPLIED MATHEMATICS, 27(3) (3), 425 - 441, 英語[査読有り]研究論文(学術雑誌)
- 2010年06月, JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 234(4) (4), 1036 - 1048, 英語[査読有り]研究論文(学術雑誌)
- 2010年06月, JOURNAL OF COMPUTATIONAL PHYSICS, 229(11) (11), 4382 - 4423, 英語[査読有り]研究論文(学術雑誌)
- 一般社団法人 日本応用数理学会, 2010年, 応用数理, 20(1) (1), 75 - 76, 日本語
- 一般社団法人 日本応用数理学会, 2009年, 応用数理, 19(3) (3), 205 - 206, 日本語
- 2009年, NUMERICAL ANALYSIS AND APPLIED MATHEMATICS, VOLS 1 AND 2, 1168, 892 - 895, 英語An Energy Conservative Numerical Scheme on Mixed Meshes for the Nonlinear Schrodinger Equation[査読有り]研究論文(国際会議プロシーディングス)
- 離散変分法は,解が保存・散逸的性質を持つ偏微分方程式に対し,元の方程式と同様に数値解が保存・散逸的性質を持つように差分スキームを導出するための方法である.離散変分法はこれまで等間隔格子上での利用が主であったが,本論文ではこれを多次元の非一様格子へと拡張する.一般社団法人 日本応用数理学会, 2009年, 応用数理学会論文誌,, 19(4) (4), 371-431 - 431, 日本語[査読有り]研究論文(学術雑誌)
- 計算機資源は有限のため,無限領域上における波動シミュレーションでは領域の打ち切りが必要となるが,このときに生じた打ち切り断面上で与える境界条件は,得られる数値解の質に大きな影響を与える,本論文では,非粘性圧縮流体の等エントロピー流れに対し,ある人工的境界条件を導出し,その境界条件のもとで解の評価を与える.さらに,得られた境界条件はよく知られたThompsonの無反射境界条件と等価であることを示す.一般社団法人 日本応用数理学会, 2008年, 日本応用数理学会論文誌, 18(3) (3), 447 - 471, 日本語
- 2006年, Journal of Computational and Applied Mathematics, 197(1) (1)研究論文(学術雑誌)
- 2016年06月, シミュレーション, 35(2) (2), 日本語微分方程式モデルによる楽器シミュレーション[招待有り]記事・総説・解説・論説等(学術雑誌)
- 京都大学, 2015年07月, 数理解析研究所講究録, 1957, 14 - 26, 日本語Webster方程式に対する離散勾配法とその力学的不変性について (新時代の科学技術を牽引する数値解析学)
- 日本数学会, 2014年, 数学, 66(1) (1), 107 - 111, 日本語書評 D. Furihata and T. Matsuo : Discrete Variational Derivative Method : A Structure-Preserving Numerical Method for Partial Differential Equations
- 京都大学, 2012年04月, 数理解析研究所講究録, 1791, 87 - 96, 日本語ある半離散スキームによるソリトンシミュレーションについて (科学技術計算における理論と応用の新展開)
- 京都大学, 2010年11月, 数理解析研究所講究録, 1719, 61 - 73, 日本語ハミルトン偏微分方程式に対する解析力学的空間離散化法とその応用 (数値解析と数値計算アルゴリズムの最近の展開)
- 波動現象シミュレーションのための無反射境界の作り方(<小特集>超ロバスト計算原理とモデリング・シミュレーション)Nonreflecting boundary conditions for numerical simulations of waves are reviewed. We describe the idea of the classical Engquist-Majda boundary condition for linear wave equations and the Hedstrom boundary condition for quasilinear hyperbolic systems. Some comments on the theoretical aspects of the boundary treatments such as the validity of the nonreflecting boundary conditions are provided. The recent developments on this subject are also discussed.日本シミュレーション学会, 2007年06月15日, シミュレーション, 26(2) (2), 84 - 89, 日本語
- 空間多次元Navier-Stokes方程式に対する無反射境界条件空力学の諸問題を扱った数値シミュレーションにおいては, 広大な現実の空間に比べ計算機の中で扱うことのできる領域は高々有限であるため, 計算対象となる空間の打ち切りが必要となる.このとき, 打ち切られた断面という人工的な境界が生じてしまうが, そのような人工的な境界上で特別な取り扱いをしなければ, 現実には存在しない反射波が生成されてしまい, 現実的な解を得ることはできない.そこで, 無反射境界条件, すなわち, 人工的な境界上で反射が起こらないようにするための境界条件の設定が重要となる.無反射境界条件は既にいくつか提案されているが, 特にPoinsot-Leleの境界条件は, その頑健性と実装の容易さから現在広く利用されている手法のひとつとなっている.しかし, Poinsot-Leleの手法の基礎となったThompsonの境界条件の有効性が理論的に保証されているのは波が境界に対して垂直に入射している場合のみである.この問題点は以前から指摘され, その改善が望まれていた.本論文ではこの要望に答える.すなわち, 数値計算時のデータを利用することによって, 流れの向きに対する仮定をおかないEuler方程式に対する無反射境界条件を提案し, そのNavier-Stokes方程式への拡張法について述べる.日本流体力学会, 2005年02月25日, ながれ : 日本流体力学会誌, 24(1) (1), 81 - 91, 日本語
- G221 特性曲線法を用いた新しい無反射境界条件(G-22 波動・音・衝撃波(2),一般講演)Because the computational resources are finite, one needs to truncate the computational domain when he/she simulates a physical problem. This truncation gives rise to non-physical artificial boundaries and one cannot obtain proper solutions without appropriate boundary conditions on such boundaries. Practically nonreflecting boundary conditions, which are boundary conditions that prevent the generation of reflections, are of great importance. Most popular methods for the Navier-Stokes equations right now are boundary conditions by Poinsot and Lele. However, their methods are based on Thompson's boundary condition for the Euler equations, which are essentially one-dimensional, and hence are valid only when the flow is perpendicular to the boundary. Here we propose a boundary condition for the Navier-Stokes equations which does not require the assumption for the direction of flow. Our basic idea is to estimate the direction of the flow with numerical data.日本流体力学会, 2004年, 日本流体力学会年会講演論文集, 2004, 456 - 457, 日本語
- 日本数学会2025年度年会, 2025年03月, 日本語, 国内会議Navier–Stokes 方程式に対する PINNs の解の誤差解析口頭発表(一般)
- 日本数学会2025年度年会, 2025年03月, 日本語, 国内会議幾何学的深層科学技術計算[招待有り]口頭発表(招待・特別)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議Modeling Coupled Systems by Neural Networks with Poisson Structures and Ports口頭発表(一般)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議Refinement of the average vector field method for Hamiltonian systems using neural networks口頭発表(一般)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議Learning Hamiltonian Partial Differential Equations Using DeepONet with a Symplectic Branch Network口頭発表(一般)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議Learning Hamiltonian Density Using DeepONet for Modeling Wave Equations口頭発表(一般)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議An Infinite Dimensional LSSL with Infinite Dimensional HiPPO口頭発表(一般)
- International Conference on Scientific Computing and Machine Learning 2025, 2025年03月, 日本語, 国内会議Energy-consistent Neural Operator Learning口頭発表(一般)
- Workshop on Dynamical Systems and Machine Learning, 2025年02月, 英語, 国際会議Model Reduction of Neural Operators by Infinite-Dimensional Singular Value Decomposition[招待有り]口頭発表(招待・特別)
- 日本応用数理学会環瀬戸内応用数理研究部会第28回シンポジウム, 2024年12月, 日本語, 国内会議Ge-Marsden の定理に基づくSympNets の改良の試み口頭発表(一般)
- RIMS研究集会「幾何構造と微分方程式 - 対称性・特異点・力学系の視点から-」, 2024年12月, 英語, 国際会議On a posteriori estimates of physics-informed neural networks for solving partial differential equations[招待有り]口頭発表(招待・特別)
- 第27回情報論的学習理論ワークショップ (IBIS2024), 2024年11月, 日本語, 国内会議波動方程式のハミルトニアン密度のDeepONetによる作用素学習口頭発表(一般)
- 第49回ASE研究会開催, 2024年10月, 日本語, 国内会議深層科学技術計算[招待有り]口頭発表(招待・特別)
- Plasma Simulator Symposium 2024, 2024年09月, 日本語, 国内会議深層科学技術計算:深層学習の物理モデリング・シミュレーションへの応用[招待有り]口頭発表(招待・特別)
- 日本応用数理学会2024年度年会, 2024年09月, 日本語, 国内会議非線形波動のモデリングのためのハミルトニアン密度の作用素学習口頭発表(一般)
- REMODEL-DSC Workshop on Machine Learning and Physics, 2024年08月, 英語, 国際会議Hyperbolic Partial Differential Equations Derived From Hippo Matricesポスター発表
- REMODEL-DSC Workshop on Machine Learning and Physics, 2024年08月, 英語, 国際会議Application of DeepONet for learning Hamiltonian PDEsポスター発表
- REMODEL-DSC Workshop on Machine Learning and Physics, 2024年08月, 英語, 国際会議Structure-preserving methods for a class of dissipative differential equations[招待有り]口頭発表(招待・特別)
- REMODEL-DSC Workshop on Structure-Preserving Numerical Methods and Machine Learning, 2024年08月, 英語, 国際会議Geometric Deep Energy-Based Models for Physics[招待有り]口頭発表(招待・特別)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2024, 2024年07月, 英語, 国際会議Neural Operators for Hamiltonian and Dissipative PDEsポスター発表
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2024, 2024年07月, 英語, 国際会議Improved estimate of the number of input points of DeepONet口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2024, 2024年07月, 英語, 国際会議Operator Learning of Hamiltonian Density for Modeling Nonlinear Waves口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2024, 2024年07月, 英語, 国際会議Enhancing Modeling Accuracy via Discriminating Hamiltonian Systems口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2024, 2024年07月, 英語, 国際会議An error bound of PINNs for solving differential equations口頭発表(一般)
- 第29回計算工学講演会, 2024年06月, 日本語, 国内会議PINNによってエネルギー保存則・エントロピー増大則を保つGENERIC系の作用素学習口頭発表(一般)
- 第29回計算工学講演会, 2024年06月, 日本語, 国内会議Physics-Informed Neural Networksの誤差解析について口頭発表(一般)
- The Data Science seminar in University of Birmingham, 2024年05月, 英語, 国際会議Deep Learning Models for Physical Modeling[招待有り]口頭発表(招待・特別)
- PhysML Workshop 2024, 2024年05月, 英語, 国際会議An error bound of physics-informed neural networks for solving differential equations[招待有り]口頭発表(招待・特別)
- The DNA (Differential Equations and Numerical Analysis) Seminar, 2024年05月, 英語, 国際会議Deep Discrete-Time Models for Physics[招待有り]口頭発表(招待・特別)
- Cambridge Image Analysis Sminar, 2024年05月, 英語, 国際会議Geometric Deep Energy-Based Models for Physics[招待有り]口頭発表(招待・特別)
- MfIP連携探索ワークショップ「数学を軸とする新たな価値創造に向けて」, 2024年04月, 日本語, 国内会議深層科学技術計算とそれを支える数学[招待有り]口頭発表(招待・特別)
- BIRS Workshop: Structured Machine Learning and Time–Stepping for Dynamical Systems, 2024年02月, 英語, 国際会議Numerical integrators for learning neural ordinary differential equation models[招待有り]口頭発表(招待・特別)
- 日本応用数理学会環瀬戸内応用数理研究部会第27回シンポジウム, 2023年12月, 日本語, 国内会議DeepONet による発展型偏微分方程式の学習口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第27回シンポジウム, 2023年12月, 日本語, 国内会議DeepONet による非線形力学系の解の予測における入力点数の評価の改良口頭発表(一般)
- IMI研究集会「新時代における高性能科学技術計算法の探究」, 2023年11月, 日本語, 国内会議深層物理モデルにおける数値解析技術の応用について[招待有り]口頭発表(招待・特別)
- RIMS研究集会「新時代における高性能科学技術計算法の探究」, 2023年10月, 日本語, 国内会議幾何学的深層学習[招待有り]口頭発表(招待・特別)
- 日本数学会2023年度秋季総合分科会, 2023年09月, 日本語, 国内会議末梢血造血幹細胞動員データ解析のためのグレブナー基底による変数分 類手法口頭発表(一般)
- 持続的環境エネルギー社会共創研究機構 研究所間交流会, 2023年09月, 日本語, 国内会議アクティブエイジングプロジェクトにおける社会ネットワーク解析[招待有り]口頭発表(招待・特別)
- 10th International Congress on Industrial and Applied Mathematics (ICIAM2023), 2023年08月, 英語, 国際会議Application of the Kernel Method to Learning Hamiltonian Equations口頭発表(一般)
- 10th International Congress on Industrial and Applied Mathematics (ICIAM2023), 2023年08月, 英語, 国際会議Structure-Preserving Learning for GENERIC systems口頭発表(一般)
- 10th International Congress on Industrial and Applied Mathematics (ICIAM2023), 2023年08月, 英語, 国際会議Geometric Integrators for Neural Symplectic Forms口頭発表(一般)
- Maths4DL Deep Learning for Computational Physics conference, 2023年07月, 英語, 国際会議Neural symplectic form and its variational principleポスター発表
- 2023年度 第37回 人工知能学会全国大会 (JSAI2023), 2023年06月, 日本語, 国内会議物理システムにおける深層学習のための損失関数口頭発表(一般)
- 第28回計算工学講演会, 2023年06月, 日本語, 国内会議カーネル法によるハミルトン系の学習と乱択化による高速化口頭発表(一般)
- 数学と諸分野の連携にむけた若手数学者交流会2023, 2023年03月, 日本語幾何学的深層科学技術計算 ~深層学習による物理モデリング・ シミュレーション~[招待有り]口頭発表(招待・特別)
- 日本応用数理学会環瀬戸内応用数理研究部会第26回シンポジウム, 2022年12月, 日本語在変数をもつハミルトニアンニューラルネットワークのハミルトン構造をもたないデータへの適用について口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第26回シンポジウム, 2022年12月, 日本語ハミルトン系に対するカーネル法によるモデリング口頭発表(一般)
- 第25回情報論的学習理論ワークショップ (IBIS2022), 2022年11月, 日本語幾何学的力学と深層学習の連携による物理現象の構造保存型モデリング[招待有り]口頭発表(招待・特別)
- 第35回計算力学講演会, 2022年11月, 日本語深層科学技術計算の最新動向 ー幾何学的深層科学技術計算ー[招待有り]口頭発表(招待・特別)
- 日本数学会2022年度秋季総合分科会, 2022年09月, 日本語神経ネットワーク動画像からのモデリングの試み口頭発表(一般)
- 日本数学会2022年度秋季総合分科会, 2022年09月, 日本語一般化 Dissipative SymODEN の GENERIC 形式口頭発表(一般)
- 日本数学会2022年度秋季総合分科会, 2022年09月, 日本語ニューラルシンプレクティック形式と変分原理の両立性について口頭発表(一般)
- 日本応用数理学会2022年度年会, 2022年09月, 日本語複数の研究分野の連携と数理科学[招待有り]口頭発表(招待・特別)
- 日本応用数理学会2022年度年会, 2022年09月, 日本語GENERICシステムに対する構造保存型深層物理モデル口頭発表(一般)
- 日本応用数理学会2022年度年会, 2022年09月, 日本語深層学習を用いてデータから力学系の第一積分を発見し保存するモデル化法口頭発表(一般)
- 日本応用数理学会2022年度年会, 2022年09月, 日本語交流アンケートデータからのネットワーク特徴量推定について口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2022, 2022年07月, 英語Learning GENERIC Systems Using Neural Symplectic Forms口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2022, 2022年07月, 英語Theoretical analysis of approximation properties of Hamiltonian neural networks口頭発表(一般)
- International Conference on Scientific Computation and Differential Equations (SciCADE) 2022, 2022年07月, 英語Neural symplectic form and coordinate-free learning of Hamiltonian dynamics口頭発表(一般)
- 電子情報通信学会 情報論的学習理論と機械学習研究会(IBISML), 2022年06月, 日本語射影法を用いて系の第一積分を発見し保存するNeural ODE口頭発表(一般)
- 2022年度 第36回人工知能学会全国大会(JSAI2022), 2022年06月, 日本語アンバランスを考慮した深層学習による物理系の学習口頭発表(一般)
- 電子情報通信学会 NOLTAソサイエティ大会, 2022年06月, 日本語Imbalance-aware lossを用いた深層学習による物理系の学習口頭発表(一般)
- 第27回計算工学講演会, 2022年06月, 日本語Neural Symplectic 形式によるGENERICシステムの学習口頭発表(一般)
- Geometric Deep Energy- Based Models for Physics, Workshop on Functional Inference and Machine Intelligence (FIMI2022), 2022, 2022年03月, 英語Geometric Deep Energy- Based Models for Physics[招待有り]口頭発表(招待・特別)
- 電子情報通信学会技術研究報告 複雑コミュニケーションサイエンス研究会(CCS), 2022年03月, 日本語Learning Physical Systems with Imbalance-Aware Deep Learning口頭発表(一般)
- 日本応用数理学会第18 回研究 部会連合発表会, 2022年03月, 日本語社会的つながりの次 数分布からの交流ネットワーク生成モデルの提案口頭発表(一般)
- 日本応用数理学会第18 回研究部会連合発表会, 2022年03月, 日本語ニューラルシンプレクティック形式とその応用口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第25 回シンポジウ ム, 2021年12月, 日本語非平衡熱力学による摩擦付き質点バネ系に対する数値解法とその刻み幅条件口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第25 回シンポジウ ム, 2021年12月, 日本語シンプレクティック形式の学習による一般座標系での 深層物理モデル口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第25 回シンポジウム, 2021年12月, 日本語ハミルトニアンニューラルネットワークの安定性について口頭発表(一般)
- 電子情報通信学会技術研究報告複雑コミュニケーションサイ エンス研究会(CCS), 2021年11月, 日本語シンプレクティック随伴変数法に基づく省メモリな Neural ODE の学習口頭発表(一般)
- 第24 回情報論的学習理論ワークショップ(IBIS2021), 2021年11月, 日本語ハミルトニアンニューラルネットワークの理論評価と KAM 理論への応用口頭発表(一般)
- 第24 回情報論的学習理論ワークショップ(IBIS2021), 2021年11月, 日本語シンプレクティック随伴変数法による高速省メモリ なNeural ODE の勾配計算口頭発表(一般)
- 第24 回情報論的学習理論ワークショップ (IBIS2021), 2021年11月, 日本語ニューラルシンプレクティック形式とそれによる一般座標系でのハミルトン方程式の学習口頭発表(一般)
- DMV-OMG Annual Conference 2021, 2021年09月, 英語Geometric Energy-Based Deep-Learning Models for Physics[招待有り]口頭発表(招待・特別)
- 第 31 回日本数理生物学会大会(2021 年度年会), 2021年09月, 日本語同定不可能モデルの解析:パラメータ多様体とその展開口頭発表(一般)
- 電子情報通信学会情報論的学習理論と機械学習研究会(IBISML), 2021年06月, 日本語シンプレクティック数値積分法を用いたNeural ODE の学習口頭発表(一般)
- 第35 回人工知能学会全国大会(JSAI2021), 2021年06月, 日本語離散時間ラグランジュ力学のニューラルネットワー クによるモデル化口頭発表(一般)
- 2021 年度第35 回人工知能学会全国大会 (JSAI2021), 2021年06月, 日本語物理現象のエネルギー挙動を離散時間で保証する深層学習シミュレーション口頭発表(一般)
- 第26回計算工学講演会, 2021年05月, 日本語⼀般座標系におけるエネルギーベース物理モデル口頭発表(一般)
- 明治大学共同利用・共同研究拠点研究集会「高度な自動運転を実現するための数理の現状と課題」, 2021年03月, 日本語, オンライン, 国内会議深層学習を用いたエネルギーベースのモデリング・シ ミュレーションフレームワーク[招待有り]口頭発表(一般)
- 日本応用数理学会第17回研究部会連合発表会, 2021年03月, 日本語, オンライン, 国内会議Koopman 作用素を利用した発展型ネットワーク予測の試み口頭発表(一般)
- 日本応用数理学会第17回研究部会連合発表会, 2021年03月, 日本語, オンライン, 国内会議アトラクターのトポロジーに着目した因果推定手法について口頭発表(一般)
- 日本応用数理学会第17回研究部会連合発表会, 2021年03月, 日本語, オンライン, 国内会議非線形状態空間システム解析における代数的マトロイドの応用について口頭発表(一般)
- Workshop: シミュレーションとモデリングのための計算代数 2021, 2021年02月, 日本語, オンライン, 国内会議深層学習によるエネルギーベース物理モデル, その2[招待有り]口頭発表(一般)
- Workshop: シミュレーションとモデリングのための計算代数 2021, 2021年02月, 日本語, オンライン, 国内会議深層学習によるエネルギーベース物理モデル, その1[招待有り]口頭発表(一般)
- 日本ディープラーニング協会主催 NeurIPS 2020 技術報告会, 2021年01月, 日本語, オンライン, 国内会議Deep Energy-Based Modeling of Discrete-Time Physics[招待有り]口頭発表(一般)
- 数値解析セミナー, 2021年01月, 日本語, オンライン, 国内会議DGNet: エネルギー保存・散逸則を保つ深層物理モデリングとそれに関する理論・応用[招待有り]口頭発表(一般)
- 2020年度応用数学合同研究集会, 2020年12月, 日本語, オンライン, 国内会議潜在変数をもつニューラル微分方程式に対する代数的考察口頭発表(一般)
- 2020年度応用数学合同研究集会, 2020年12月, 日本語, オンライン, 国内会議深層フェーズフィールドモデリング口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第24回シンポジウム, 2020年12月, 日本語, オンライン, 国内会議自然系の連成とシンプレクティック形式口頭発表(一般)
- NeurIPS2020 Workshop on Machine Learning and the Physical Sciences (ML4PS), 2020年12月, 英語, オンライン, 国内会議The Error Analysis of Numerical Integrators for Deep Neural Network Modeling of Differential Equationsポスター発表
- Establishing International Research Network of Mathematical Oncology (Fusion of Mathematics and Biology), 2020年10月, 英語, 大阪, 国内会議The parameter variety of unidentifiable state-space models and its applications to analysis of biological systems口頭発表(一般)
- 日本応用数理学会2020年度年会, 2020年09月, 日本語, オンライン, 国内会議分布系のカオス同期化とニューラルネットワークを用いた秘匿通信システム口頭発表(一般)
- 日本応用数理学会2020年度年会, 2020年09月, 日本語, オンライン, 国内会議常微分方程式モデルの学習における離散化手法の影響について口頭発表(一般)
- 第25回計算工学講演会, 2020年06月, 日本語, オンライン, 国内会議ピアノの弦と駒の連成シミュレーションによるエネルギー移動の可視化口頭発表(一般)
- 日本応用数理学会第16回研究部会連合発表会, 2020年03月, 東京, 国内会議波動方程式と弾性方程式からなる連成系のシンプレクティッ ク性について口頭発表(一般)
- 日本応用数理学会第16回研究部会連合発表会, 2020年03月, 東京, 国内会議時間方向対称性を利用した2つのエネルギー保存数値解法の等価条件について口頭発表(一般)
- 日本応用数理学会第16回研究部会連合発表会, 2020年03月, 東京, 国内会議微分代数の応用に向けた多項式常微分方程式モデルの簡約口頭発表(一般)
- Workshop: シミュレーションとモデリングのための計算代数 2020, 2020年01月, 神戸, 国内会議微分代数に基づく数理モデリングアプローチ口頭発表(一般)
- Workshop: シミュレーションとモデリングのための計算代数 2020, 2020年01月, 神戸, 国内会議幾何学的離散力学と対称性 II[招待有り]口頭発表(一般)
- Workshop: シミュレーションとモデリングのための計算代数 2020, 2020年01月, 神戸, 国内会議幾何学的離散力学と対称性 I[招待有り]口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第23回シンポジウム, 2019年12月, 神戸, 国内会議指数型分布族の定める多様体上の離散力学に基づく時系列モデルとネットワーク解析への応用口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第23回シンポジウム, 2019年12月, 神戸, 国内会議高頻度データに対する再帰型ニューラルネットモデルとその比較口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第23回シンポジウム, 2019年12月, 神戸, 国内会議波動方程式と弾性方程式の構造保存型連成数値計算口頭発表(一般)
- 2019 年度応用数学合同研究集会, 2019年12月, 滋賀, 国内会議同定不可能モデルに対するパラメータ多様体による解析とその近似導出について口頭発表(一般)
- 2019 年度応用数学合同研究集会, 2019年12月, 滋賀, 国内会議自動微分による離散力学とアルゴリズム的数値解析口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第22回シンポジウム, 2018年12月, 日本語, 香川, 国内会議微分代数方程式モデルのモデルパラメータと解に関するグレブナー基底を用いた解析口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第22回シンポジウム, 2018年12月, 日本語, 香川, 国内会議波動型偏微分方程式に対する幾何学的弱形式口頭発表(一般)
- RIMS研究集会, 2018年11月, 日本語, 京都, 国内会議アレルギー疾患の個別化医療に向けた抗原・抗体の体内動態シミュレーション口頭発表(一般)
- 情報計算科学生物学会2018年大会, 2018年10月, 日本語, 東京, 国内会議Modeling and simulations of the kinetics of antigens and antibodies towards personalized medicine for allergies口頭発表(一般)
- 日本応用数理学会2018年度年会, 2018年09月, 日本語, 愛知, 国内会議統計多様体上の状態空間モデルを用いた発展型ネットワーク解析口頭発表(一般)
- 日本応用数理学会2018年度年会, 2018年09月, 日本語, 愛知, 国内会議抗原・抗体の体内動態の定量的解析に向けたモデルパラメータの多様性に対する考察口頭発表(一般)
- 日本応用数理学会2018年度年会, 2018年09月, 日本語, 愛知, 国内会議アンケートデータを用いた交流ネットワーク推定手法口頭発表(一般)
- MIMS現象数理学研究拠点共同研究集会「幾何的解析と形状表現の数理」, 2018年08月, 日本語, 東京, 国内会議情報幾何学を用いた発展型ネットワークモデルに基づく相転移に着目した異常検知の試み口頭発表(一般)
- SIAM Conference on the Life Science (LS18), 2018年08月, 英語, Minneapolis, 国内会議Parameters of Models using Dynamical Systems with Conservation Laws口頭発表(一般)
- SIAM Conference on the Life Science (LS18), 2018年08月, 英語, Minneapolis, 国内会議Modeling the Kinetics of Antigens and Antibodies for Analysis of the Mechanism of Allergy口頭発表(一般)
- Data Science, Statistics & Visualisation (DSSV 2018), 2018年07月, 英語, Wien, 国内会議Parameter estimation for compartment models of biological systems口頭発表(一般)
- Data Science, Statistics & Visualisation (DSSV 2018), 2018年07月, 英語, Wien, 国内会議Autoregressive models on statistical Riemannian manifolds for analysis of evolutionary networks口頭発表(一般)
- The 13th World Congress in Computational Mechanics, 2018年07月, 英語, New York, 国内会議Application of Hamiltonian Flows to Exploring Parameters of Mathematical Models in Situations with Insufficient Data口頭発表(一般)
- 第47回数値解析シンポジウム, 2018年06月, 日本語, 福井, 国内会議体内動態に対するコンパートメントモデルのモデルパラメータ推定手法について口頭発表(一般)
- 第47回数値解析シンポジウム, 2018年06月, 日本語, 福井, 国内会議潜在変数ネットワークモデルを用いた放牧牛の交流ネットワーク解析口頭発表(一般)
- 第47回数値解析シンポジウム, 2018年06月, 日本語, 福井, 国内会議あるテーマパークにおける地形的集客効果の感度分析口頭発表(一般)
- 日本応用数理学会第14回研究部会連合発表会, 2018年03月, 日本語, 国内会議有限要素外積解析に対するRRGMRES法口頭発表(一般)
- 日本数学会2018年度年会, 2018年03月, 日本語, 国内会議変分原理に基づくエネルギー保存数値解法の Lie 群上への拡張口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第21回シンポジウム, 2018年03月, 日本語, 国内会議統計多様体上のARモデルを用いた発展型ネットワーク解析口頭発表(一般)
- 応用数理 学生・若手研究者のための研究交流会, 2018年03月, 日本語, 国内会議質点ばね系を用いたレザバーコンピューティングの数値実験口頭発表(一般)
- 日本応用数理学会第14回研究部会連合発表会, 2018年03月, 日本語, 国内会議アレルギー発症メカニズムの解析に向けた抗原・抗体の体内動態モデルの構築, 及び, Husbyらの実験データに対するパラメータ推定とその考察口頭発表(一般)
- 日本応用数理学会環瀬戸内応用数理研究部会第21回シンポジウム, 2018年03月, 日本語, 国内会議アレルギー発症シミュレーションに向けた生理学的薬物動態モデルの応用口頭発表(一般)
- SIAM Conference on Parallel Processing for Scientific Computing, 2018年03月, 英語, 国際会議Energy-Preserving Parareal Algorithm for the Hamilton Equationシンポジウム・ワークショップパネル(指名)
- 第46回数値解析シンポジウム, 2017年, 日本語, 国内会議離散偏導関数法と数値積分の併用口頭発表(一般)
- 第46回数値解析シンポジウム, 2017年, 日本語, 国内会議離散外積解析における離散 Hodge スター作用素の誤差評価口頭発表(一般)
- 日本応用数理学会2017年度年会, 2017年, 日本語, 国内会議離散外積解析から導かれる有限積分法のマルチシンプレクティック性について口頭発表(一般)
- 第46回数値解析シンポジウム, 2017年, 日本語, 国内会議速度比例減衰項をもつ系に対する変分原理を利用した数値解法とその比較口頭発表(一般)
- 日本応用数理学会2017年度年会, 2017年, 日本語, 国内会議指数ランダムグラフモデルに基づくネットワークに対するARモデルポスター発表
- the International Conference on Scientific Computation And Differential Equations 2017 (SciCADE 2017), 2017年, 英語, 国際会議Regression model on statistical manifolds and its application to evolutionary network analysis口頭発表(一般)
- the International Conference on Scientific Computation And Differential Equations 2017 (SciCADE 2017), 2017年, 英語, 国際会議Discrete partial derivative method with numerical integrations口頭発表(一般)
- 2017年度応用数学合同研究集会, 2017年, 日本語, 国内会議curl-curl型偏微分方程式に対する有限要素外積解析の応用口頭発表(一般)
- the International Conference on Scientific Computation And Differential Equations 2017 (SciCADE 2017), 2017年, 英語, 国際会議Automatic discrete differentiation and its applications口頭発表(一般)
- 日本応用数理学会2016年度年会, 2016年09月, 日本語, 国内会議変分原理に基づくエネルギー保存数値解法の一般のHamilton系への拡張口頭発表(一般)
- 研究集会「常微分方程式の数値解法とその周辺2016」, 2016年07月, 日本語, 国内会議離散化した heavy-ball-with-friction method のパラメータについて口頭発表(一般)
- 第45回数値解析シンポジウム, 2016年06月, 日本語, 国内会議波動方程式に対するシンプレクティックかつエネルギー保存スキームについて口頭発表(一般)
- 第45回数値解析シンポジウム, 2016年06月, 日本語, 国内会議散逸型偏微分方程式に対するある種の変分原理に基づく散逸スキームの導出法口頭発表(一般)
- 第45回数値解析シンポジウム, 2016年06月, 日本語, 国内会議曲面上の熱方程式に対する散逸性保存型数値解法の導出と評価口頭発表(一般)
- 第45回数値解析シンポジウム, 2016年06月, 日本語, 国内会議Webster方程式に対するある数値解法の長時間挙動について口頭発表(一般)
- 日本応用数理学会第12回研究部会連合発表会, 2016年03月, 日本語, 神戸学院大学, 国内会議地域コミュニティ構造の変化と改善に対する統計解析手法口頭発表(一般)
- 日本応用数理学会研究部会連合発表会, 2016年03月, 日本語, 神戸学院大学, 国内会議自動離散微分とその応用口頭発表(一般)
- 日本数学会2016年度年会, 2016年03月, 日本語, 筑波大学, 国内会議散逸型構造保存型数値解法の多層パーセプトロン学習法への応用口頭発表(一般)
- 応用数学合同研究集会, 2015年12月, 日本語, 龍谷大学, 国内会議地域コミュニティの構造変化に対する検定理論口頭発表(一般)
- 研究会「数理構造保存を接点とした数学・HPC・実科学のクロスオーバー」, 2015年12月, 日本語, 電気通信大学, 国内会議Caldirola-Kanai型変分原理に基づく構造保存型数値解法と多層パーセプトロン学習法への応用について口頭発表(一般)
- 日本応用数理学会2015年度年会, 2015年09月, 日本語, 金沢大学, 国内会議対称性を利用した離散勾配法におけるLegendre変換に関する考察[招待有り]口頭発表(招待・特別)
- 日本応用数理学会2015年度年会, 2015年09月, 日本語, 金沢大学, 国内会議ハミルトン方程式に対する時間対称性を用いた離散勾配スキームの導出法口頭発表(一般)
- 日本応用数理学会2015年度年会, 2015年09月, 日本語, 金沢大学, 国内会議シンプレクティック数値積分法による力学的摂動口頭発表(一般)
- 日本応用数理学会2015年度年会, 2015年09月, 日本語, 金沢大学, 国内会議ある種の散逸型微分方程式に対する構造保存型数値解法口頭発表(一般)
- the International Conference on Scientific Computation And Differential Equations 2015 (SciCADE 2015), 2015年09月, 英語, University of Potsdam, 国際会議Structure-preserving method for a certain class of dissipative differential equations口頭発表(一般)
- the International Conference on Scientific Computation And Differential Equations 2015 (SciCADE 2015), 2015年09月, 英語, University of Potsdam, 国際会議Energy-preserving discrete gradient schemes for the Hamilton equation based on the variational principle口頭発表(一般)
- Computational and Geometric Approaches for Nonlinear Phenomena, 2015年08月, 英語, 早稲田大学, 国際会議Numerical integrations that preserve energy behaviors using the variational principle口頭発表(一般)
- Recent developments in numerical analysis with special emphasis on complex analysis, 2015年07月, 英語, 東京大学, 国際会議Structure-preserving numerical integrators for the KdV equation using an almost complex structure口頭発表(一般)
- 日本老年社会科学会第57回大会, 2015年06月, 日本語, 国内会議地域高齢者を対象とした健康教室による参加者間交流ネットワーク形成に関する研究ポスター発表
- 第44回数値解析シンポジウム, 2015年06月, 日本語, ぶどうの丘, 国内会議大規模ネットワークにおける複数ノード組に対する重要度の特徴付け口頭発表(一般)
- 第44回数値解析シンポジウム, 2015年06月, 日本語, ぶどうの丘, 国内会議ピアノの物理モデルとその効率的な数値計算法の検討口頭発表(一般)
- 第44回数値解析シンポジウム, 2015年06月, 日本語, ぶどうの丘, 国内会議L2射影を用いた離散偏導関数法による弦のサウンドレンダリング口頭発表(一般)
- 日本応用数理学会研究部会連合発表会, 2015年03月, 日本語, 東京, 国内会議測地線方程式に対する離散勾配法の適用とアインシュタイン方程式の数値解を用いるための基礎検討口頭発表(一般)
- 日本数学会2014年度年会, 2014年03月, 日本語, 東京, 国内会議境界付き多様体上における有限要素外積解析の弱形式の適切性について口頭発表(一般)
- 研究集会「常微分方程式の数値解法とその周辺2014」, 2014年03月, 日本語, 静岡, 国内会議楽器シミュレーションに対する構造保存型数値解法の応用と関連する数理的課題口頭発表(一般)
- RIMS研究集会「新時代の科学技術を牽引する数値解析学」, 2014年, 日本語, 京都, 国内会議離散勾配法のRiemann構造不変性とシンプレクティック幾何学的再構築[招待有り]口頭発表(招待・特別)
- 応用数学合同研究集会, 2014年, 日本語, 滋賀, 国内会議数値相対論のための測地線方程式に対する構造保存型数値解法の適用口頭発表(一般)
- 第3回岐阜数理科学研究会, 2014年, 日本語, 岐阜, 国内会議幾何学的構造保存型数値解法に対する力学理論的アプローチ[招待有り]口頭発表(招待・特別)
- 第43回数値解析シンポジウム, 2014年, 日本語, 沖縄, 国内会議異なる内積により得られる Webster 方程式の2つのハミルトン構造口頭発表(一般)
- 日本応用数理学会2014年度年会, 2014年, 日本語, 東京, 国内会議異なるRiemann構造をもつWebster方程式に対する離散変分導関数法の不変性口頭発表(一般)
- 日本学術会議第4回計算力学シンポジウム, 2014年, 日本語, 東京, 国内会議ハミルトン偏微分方程式に対する構造保存型数値解法[招待有り]口頭発表(招待・特別)
- 第43回数値解析シンポジウム, 2014年, 日本語, 沖縄, 国内会議シンプレクティック法による摂動を用いた太陽系の安定性検証口頭発表(一般)
- 応用数学合同研究集会, 2014年, 日本語, 滋賀, 国内会議シンプレクティック空間上の離散勾配法口頭発表(一般)
- 日本応用数理学会2014年度年会, 2014年, 日本語, 東京, 国内会議グラフに対するOllivier-Ricci曲率の数値計算口頭発表(一般)
- Foundations of Computational Mathematics Conference 2014, 2014年, 英語, ウルグアイ, 国際会議Simulation of Wind Instruments and a Geometric Invariance of the Discrete Gradient Method[招待有り]口頭発表(招待・特別)
- 流体方程式の構造と特異性に迫る数値解析・数値計算, 2014年, 英語, 愛知, 国際会議On the well-posedness of the weak form of the finite element exterior calculus on manifolds[招待有り]口頭発表(招待・特別)
- 2nd International Workshop on Numerical Linear Algebra and Its Applications, 2014年, 英語, 中国, 国際会議Application of Structure-Preserving Numerical Methods to Simulation of Musical Instruments[招待有り]口頭発表(招待・特別)
- 日本数学会 秋季総合分科会, 2013年09月, 日本語, 愛媛, 国内会議有限要素外積解析に基づく波動型方程式に対するエネルギー保存型数値解法[招待有り]口頭発表(招待・特別)
- 日本応用数理学会 2013 年度年会, 2013年09月, 日本語, 福岡, 国内会議ホロノーム拘束をもつハミルトン系に対する離散勾配法口頭発表(一般)
- 日本応用数理学会 2013 年度年会, 2013年09月, 日本語, 福岡, 国内会議シンプレクティック数値積分法における修正ハミルトニアンの存在定理について口頭発表(一般)
- the International Conference on Scientific Computation And Differential Equations 2013 (SciCADE 2013), 2013年09月, 英語, Valladolid, Spain, 国際会議Lagrangian approach of the discrete gradient method based on finite element methods口頭発表(一般)
- ワークショップ「有限体積法の数学的基盤理論の確立III」, 2013年08月, 日本語, 愛媛, 国内会議シンプレクティックフローとしてのシンプレクティック数値積分法口頭発表(一般)
- 2013 Tokyo Workshop on Structure-Preserving Methods, 2013年01月, 英語, Tokyo, 国際会議On the finite element exterior calculus for parabolic equations口頭発表(一般)
- 応用数学合同研究集会, 2012年12月, 日本語, 滋賀, 国内会議放物型方程式に対する有限要素外積解析の誤差評価について口頭発表(一般)
- 10th International Conference of Numerical Analysis and Applied Mathematics, 2012年09月, 英語, Greece, 国際会議Application of the Lagrangian Approach of the Discrete Gradient Method to Scleronomic Holonomic Systems口頭発表(一般)
- 日本応用数理学会 2012年度年会, 2012年08月, 日本語, 北海道, 国内会議ホロノミック系に対するラグランジュ力学的離散勾配法口頭発表(一般)
- 15th International Congress on Computational and Applied Mathematics, 2012年07月, 英語, Gent, Belgium, 国際会議A Lagrangian Approach to Deriving Local-Energy-Preserving Numerical Schemes for the Euler-Lagrange Partial Differential Equations口頭発表(一般)
- 有限体積法の数学的基盤理論の確立II, 2012年03月, 日本語, 福岡, 国内会議ラグランジュ力学に基づく保存型数値解法導出法とその応用口頭発表(一般)
- 常微分方程式の数値解法とその周辺 2012, 2012年03月, 日本語, 静岡, 国内会議Newton法の Parareal Algorithm による並列化口頭発表(一般)
- 日本応用数理学会研究部会連合発表会, 2012年03月, 日本語, 福岡, 国内会議Euler-Lagrange 偏微分方程式に対する局所エネルギー保存スキーム導出法口頭発表(一般)
- 2012 Tokyo Workshop on Structure-Preserving Methods, 2012年01月, 英語, Tokyo, 国際会議Backward Error Analysis of the Scheme for the KdV Equation by the Discrete Variational Derivative Method口頭発表(一般)
- 応用数学合同研究集会, 2011年12月, 日本語, 瀬田, 国内会議KdV 方程式に対するある半離散スキームの後退誤差解析口頭発表(一般)
- 日本応用数理学会「行列・固有値問題の解法とその応用」研究部会第12回研究会, 2011年11月, 日本語, 日本応用数理学会, 東京, 国内会議時間依存固有値問題の数値解法に関する基礎検討口頭発表(一般)
- RIMS研究集会「科学技術計算における理論と応用の新展開」, 2011年10月, 日本語, 京都, 国内会議ある半離散スキームによるソリトンのシミュレーションについてその他
- 日本応用数理学会 2011年度年会, 2011年09月, 日本語, 京都, 国内会議変分構造をもつ楕円型方程式に対する離散勾配法の応用口頭発表(一般)
- International Workshop on Numerical Linear Algebra and Its Applications, 2011年07月, 英語, China, 国際会議The Discrete Variational Derivative Method Based on Discrete Differential Forms口頭発表(招待・特別)
- the International Conference on Scientific Computation And Differential Equations 2011 (SciCADE 2011), 2011年07月, 英語, Canada, 国際会議A Lagrangian Approach to Deriving Energy-Preserving Numerical Schemes for the Euler-Lagrange Partial Differential Equations and Its Applications[招待有り]口頭発表(招待・特別)
- Institute of Electrical and Electronics Engineers2020年05月 - 現在
- 情報処理学会2020年02月 - 現在
- Society for Industrial and Applied Mathematics2017年01月 - 現在
- American Institute of Aeronautics and Astronautics
- 日本流体力学会
- Mathematical Association of America
- 日本数学会
- 日本応用数理学会
- 科学技術振興機構, 先端国際共同研究推進事業(ASPIRE), 次世代のためのASPIRE, 神戸大学, 2024年01月 - 2027年03月深層科学技術計算:数理科学を基盤とする物理構造と深層学習の融合
- 日本学術振興会, 科学研究費助成事業, 基盤研究(B), 神戸大学, 2022年04月01日 - 2025年03月31日感情調整と信頼の加齢変化と社会的つながりに関する縦断研究
- 国立研究開発法人科学技術振興機構, 戦略的創造研究推進事業(CREST), 2019年10月 - 2025年03月, 研究代表者幾何学的離散力学を核とする構造保存的システムモデリング・シミュレーション基盤
- 日本学術振興会, 科学研究費助成事業 基盤研究(C), 基盤研究(C), 神戸大学, 2020年04月01日 - 2024年03月31日ブラックボックス微分方程式モデルに対する保存則抽出手法とネットワーク解析への応用2021年度は,主に,与えられた時系列データの背後に潜むシンプレクティック構造を抽出する手法の開発に取り組んだ.実際の問題に現れる,保存量をもつ微分方程式の多くはハミルトン方程式であるが,ハミルトン方程式はシンプレクティック多様体上で,エネルギー関数が定めるフローとして定義される.シンプレクティック多様体は,シンプレクティック形式と呼ばれる微分2形式をもつ多様体であるが,これは,一般には,状態変数に依存してよい量であり,データから学習することが必要である. 一般に,微分2形式は歪対称行列に対応するため,素朴な手法としては,データから歪対称行列を学習する手法が考えられる.しかし,実際には,シンプレクティック形式は閉形式である必要もあり,単に歪対称行列を学習するだけでは,シンプレクティック形式に対応するとは限らない. 本研究では,de Rhamコホモロジーを考慮すると,多くのシンプレクティック多様体上で,シンプレクティック形式が微分1形式の外微分によって導かれることに着目した.具体的には,微分2形式を直接学習するのではなく,それを導く微分1形式をデータから学習することで,シンプレクティック形式以外に対応しない歪対称行列が学習されることを防ぐ手法を構築した. この手法を用いれば,与えられたデータに隠されたシンプレクティック構造を抽出することが可能となり,隠された運動方程式を発見することが出来るようになる.また,提案手法は,ハミルトン方程式の幾何学的な性質,特に座標変換不変性を利用しており,データがどのように表現されていたとしても,方程式を学習することが可能である.そのため,データの前処理とも相性が良く,この性質は,今後,様々な形で応用できる可能性がある.
- 科学研究費補助金/挑戦的研究(開拓), 2018年06月 - 2021年03月競争的資金
- 国立研究開発法人科学技術振興機構, 戦略的創造研究推進事業(さきがけ), 2016年10月 - 2020年03月, 研究代表者情報幾何学と離散力学の融合と社会ネットワーク解析への応用競争的資金
- 学術研究助成基金助成金/基盤研究(C), 2014年04月 - 2019年03月, 研究代表者競争的資金
- 科学研究費補助金/基盤研究(B), 2015年04月 - 2018年03月競争的資金
- 日本学術振興会, 科学研究費助成事業 基盤研究(B), 基盤研究(B), 東京大学, 2011年04月01日 - 2015年03月31日本研究プロジェクトでは,構造保存型の数値解法として理工学各分野で広く応用されている有限体積法に対する数学的な基盤理論の開発とその現実問題への応用を行なった。基礎的な面では、離散ソボレフの不等式、補間誤差不等式の最良定数、離散Rellichの定理、離散最大値の定理、離散微分形式などについて応用指向の進んだ結果を得ることができた。応用面では、細胞性粘菌の数理モデルに対して、構造保存型の有限体積法を開発し、いままで未解決だった離散エネルギー不等式の証明に成功した。また、離散微分形式の応用としてLagrange力学に基づくエネルギー保存型数値解法の有限体積法への拡張を行なった。
- 科学研究費補助金/若手研究(B), 2011年, 研究代表者競争的資金
- 日本学術振興会, 科学研究費助成事業 若手研究(B), 若手研究(B), 東京大学, 2007年 - 2009年圧縮流体シミュレーションで重要となる無反射境界条件について、Riemann不変量多様体に基づいた無反射境界条件を導出した。また、得られた方法は安定性に問題があったため、それを改善する修正を行ったところThompsonの無反射境界条件と呼ばれる既存の方法に一致し、その結果、Thompsonの境界条件に関する知見を得た。また、安定な実装法の開発を目指して離散変分法についての研究も行った。その結果、離散変分法をいくつかの点で拡張することに成功した。
研究シーズ
■ 研究シーズ- エネルギー挙動を保つ深層物理モデリング・シミュレーション基盤シーズカテゴリ:自然科学一般研究キーワード:ハミルトン力学, フェーズフィールドモデル, 深層学習, 物理シミュレーション研究の背景と目的:近年、深層学習により観測データから物理現象をモデル化する手法が注目されています。本研究では、古典力学やフェーズフィールドモデルなど、エネルギー関数を用いた理論でモデル化できる現象に着目し、エネルギーの保存 ・散逸則を厳密に保つ深層物理モデルを構築しています。このような手法を構築することにより、これまで物理法則が不明でシミュレーションが出来なかった現象もシミュレーションによる予測が可能になります。研究内容:本研究では、深層学習と数理物理学を連携させることで、観測されたデータから現象を記述する支配方程式を抽出する方法を開発しています。特にエネルギーを用いて記述出来る現象に着目し、新たに開発した離散版の自動微分アルゴリズムや力学理論の背景にあるシンプレクティック幾何学と組み合わせることで、様々なデータから、エネルギーの保存・散逸則などといった物理法則を保ったモデル・シミュレーションを可能にしています.期待される効果や応用分野:本研究によって、支配方程式が未知である現象に対しては新たにそのシミュレーションが可能となり、既知である現象に対しては個別の状況に合わせた精密な予測が可能となると期待されます。学習結果のモデルは、そのまま数値シミュレーション用のプログラムとして利用可能で、その計算はGPUによって並列化され、高速です。また、エネルギー保存・散逸則を厳密に保つ数値計算法を自動導出する方法としても利用できます。関係する業績:T. Matsubara, A. Ishikawa, T. Yaguchi, Deep Energy-based Modeling of Discrete-Time Physics, Advances in Neural Information Processing Systems 33 (NeurIPS 2020)