Details of the Researcher

PHOTO

Yuki Sughiyama
Section
Graduate School of Information Sciences
Job title
Specially Appointed Associate Professor(Research)
Degree
e-Rad No.
90756389

Professional Memberships 1

  • THE PHYSICAL SOCIETY OF JAPAN

Research Interests 7

  • 機械学習

  • stochastic process

  • Large deviation theory

  • Evolutionary biology

  • Biophysics

  • Thermodynamics

  • Statistical physics

Research Areas 2

  • Natural sciences / Bio-, chemical, and soft-matter physics /

  • Natural sciences / Mathematical physics and basic theory / Nonequilibrium statistical physics

Papers 36

  1. Schrödinger bridge-type diffusion models as an extension of variational autoencoders

    Kentaro Kaba, Reo Shimizu, Masayuki Ohzeki, Yuki Sughiyama

    Physical Review Research　7　(3)　2025/09/03

    Publisher: American Physical Society (APS)

    DOI： 10.1103/dxp7-4hby 　

    eISSN： 2643-1564

    More details Close

    Generative diffusion models are described by time-forward and -backward stochastic differential equations to connect the data and prior distributions. While conventional diffusion models (e.g., score-based models) only learn the backward process, more flexible frameworks have been proposed to also learn the forward process by employing the Schrödinger bridge (SB). However, due to the complexity of the mathematical structure behind SB-type models, we cannot easily give an intuitive understanding of their objective function. In this work, we propose a unified framework to construct diffusion models by reinterpreting the SB-type models as an extension of variational autoencoders. In this context, the data processing inequality plays a crucial role. As a result, we find that the objective function consists of the prior loss and drift matching parts, which enable us to reduce the numerical cost of training the forward process. Furthermore, we discuss the overfitting problem in the SB-type models in this framework.

  2. Chemical thermodynamics of growing autocatalytic processes Peer-reviewed

    Atsushi Kamimura, Yuki Sughiyama, Tetsuya J Kobayashi

    JSAP Review　93　(11)　624-629　2024/11

    DOI： 10.11470/oubutsu.93.11_624 　

  3. Thermodynamic and stoichiometric laws ruling the fates of growing systems Peer-reviewed

    Atsushi Kamimura, Yuki Sughiyama, Tetsuya J. Kobayashi

    Physical Review Research　6　(2)　2024/05/16

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.6.023173 　

    eISSN： 2643-1564

    More details Close

    We delve into growing open chemical reaction systems (CRSs) characterized by autocatalytic reactions within a variable volume, which changes in response to these reactions. Understanding the thermodynamics of such systems is crucial for comprehending biological cells and constructing protocells because it sheds light on the physical conditions necessary for their self-replication. Building on our recent work, where we developed a thermodynamic theory for growing CRSs featuring basic autocatalytic motifs with regular stoichiometric matrices, we now expand this theory to include scenarios where the stoichiometric matrix has a nontrivial left kernel space. This extension introduces conservation laws, which limit the variations in chemical species due to reactions, thereby confining the system's possible states to those compatible with its initial conditions. By considering both thermodynamic and stoichiometric constraints, we clarify the environmental and initial conditions that dictate the CRSs' fate—whether they grow, shrink, or reach equilibrium. We also find that the conserved quantities significantly influence the equilibrium state achieved by a growing CRS. These results are derived independently of specific thermodynamic potentials or reaction kinetics, therefore underscoring the fundamental impact of conservation laws on the growth of the system. Published by the American Physical Society2024

  4. Information geometry of dynamics on graphs and hypergraphs Peer-reviewed

    Tetsuya J. Kobayashi, Dimitri Loutchko, Atsushi Kamimura, Shuhei A. Horiguchi, Yuki Sughiyama

    Information Geometry　2023/12/22

    Publisher: Springer Science and Business Media LLC

    DOI： 10.1007/s41884-023-00125-w 　

    ISSN： 2511-2481

    eISSN： 2511-249X

    More details Close

    Abstract We introduce a new information-geometric structure associated with the dynamics on discrete objects such as graphs and hypergraphs. The presented setup consists of two dually flat structures built on the vertex and edge spaces, respectively. The former is the conventional duality between density and potential, e.g., the probability density and its logarithmic form induced by a convex thermodynamic function. The latter is the duality between flux and force induced by a convex and symmetric dissipation function, which drives the dynamics of the density. These two are connected topologically by the homological algebraic relation induced by the underlying discrete objects. The generalized gradient flow in this doubly dual flat structure is an extension of the gradient flows on Riemannian manifolds, which include Markov jump processes and nonlinear chemical reaction dynamics as well as the natural gradient. The information-geometric projections on this doubly dual flat structure lead to information-geometric extensions of the Helmholtz–Hodge decomposition and the Otto structure in $$L^{2}$$-Wasserstein geometry. The structure can be extended to non-gradient nonequilibrium flows, from which we also obtain the induced dually flat structure on cycle spaces. This abstract but general framework can broaden the applicability of information geometry to various problems of linear and nonlinear dynamics.

  5. Riemannian geometry of optimal driving and thermodynamic length and its application to chemical reaction networks Peer-reviewed

    Dimitri Loutchko, Yuki Sughiyama, Tetsuya J. Kobayashi

    Physical Review Research　4　(4)　2022/10/20

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.4.043049 　

    eISSN： 2643-1564

  6. Hessian geometry of nonequilibrium chemical reaction networks and entropy production decompositions Peer-reviewed

    Tetsuya J. Kobayashi, Dimitri Loutchko, Atsushi Kamimura, Yuki Sughiyama

    Physical Review Research　4　(3)　2022/09/15

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.4.033208 　

    eISSN： 2643-1564

  7. Chemical thermodynamics for growing systems Peer-reviewed

    Yuki Sughiyama, Atsushi Kamimura, Dimitri Loutchko, Tetsuya J. Kobayashi

    Physical Review Research　4　(3)　2022/09/09

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.4.033191 　

    eISSN： 2643-1564

  8. Kinetic derivation of the Hessian geometric structure in chemical reaction networks Peer-reviewed

    Tetsuya J. Kobayashi, Dimitri Loutchko, Atsushi Kamimura, Yuki Sughiyama

    Physical Review Research　4　(3)　2022/07/21

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.4.033066 　

    eISSN： 2643-1564

  9. Hessian geometric structure of chemical thermodynamic systems with stoichiometric constraints Peer-reviewed

    Yuki Sughiyama, Dimitri Loutchko, Atsushi Kamimura, Tetsuya J. Kobayashi

    Physical Review Research　4　(3)　2022/07/21

    Publisher: American Physical Society (APS)

    DOI： 10.1103/physrevresearch.4.033065 　

    eISSN： 2643-1564

  10. Lineage EM Algorithm for Inferring Latent States from Cellular Lineage Trees. International-journal Peer-reviewed

    So Nakashima, Yuki Sughiyama, Tetsuya J Kobayashi

    Bioinformatics (Oxford, England)　2020/01/23

    DOI： 10.1093/bioinformatics/btaa040 　

    More details Close

    Phenotypic variability in a population of cells can work as the bet-hedging of the cells under an unpredictably changing environment, the typical example of which is the bacterial persistence. To understand the strategy to control such phenomena, it is indispensable to identify the phenotype of each cell and its inheritance. Although recent advancements in microfluidic technology offer us useful lineage data, they are insufficient to directly identify the phenotypes of the cells. An alternative approach is to infer the phenotype from the lineage data by latent-variable estimation. To this end, however, we must resolve the bias problem in the inference from lineage called survivorship bias. In this work, we clarify how the survivor bias distorts statistical estimations. We then propose a latent-variable estimation algorithm without the survivorship bias from lineage trees based on an expectation-maximization (EM) algorithm, which we call Lineage EM algorithm (LEM). LEM provides a statistical method to identify the traits of the cells applicable to various kinds of lineage data.

  11. Fitness Gain of Individually Sensed Information by Cells Peer-reviewed

    Kobayashi, Tetsuya J., Sughiyama, Yuki

    ENTROPY　21　(10)　2019/10

    DOI： 10.3390/e21101002 　

    eISSN： 1099-4300

  12. Fitness response relation of a multi-type age-structured population dynamics Peer-reviewed

    Yuki Sughiyama, So Nakashima, Tetsuya J Kobayashi

    Phys. Rev. E　99　012413　2019

  13. The explicit form of the rate function for semi-Markov processes and its contractions Peer-reviewed

    Yuki Sughiyama, Tetsuya J Kobayashi

    J. Phys. A: Math. Theor.　51　125001　2018

  14. A Quantum Extension of Variational Bayes Inference Peer-reviewed

    Hideyuki Miyahara, Yuki Sughiyama

    Phys. Rev. A　98　(2)　022330　2018

    DOI： 10.1103/PhysRevA.98.022330 　

    ISSN： 2469-9926

    eISSN： 2469-9934

  15. Deterministic quantum annealing expectation-maximization algorithm Peer-reviewed

    Hideyuki Miyahara, Koji Tsumura, Yuki Sughiyama

    JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT　P113404　2017/11

    DOI： 10.1088/1742-5468/aa967e 　

    ISSN： 1742-5468

  16. Stochastic and information-thermodynamic structures of population dynamics in a fluctuating environment Peer-reviewed

    Tetsuya J. Kobayashi, Yuki Sughiyama

    PHYSICAL REVIEW E　96　(1)　012402　2017/07

    DOI： 10.1103/PhysRevE.96.012402 　

    ISSN： 2470-0045

    eISSN： 2470-0053

  17. 増殖・進化適応系に内在する情報論的構造 Peer-reviewed

    小林 徹也, 杉山 友規

    生物物理　57　(6)　287-290　2017

  18. Steady-state thermodynamics for population growth in fluctuating environments Peer-reviewed

    Yuki Sughiyama, Tetsuya J. Kobayashi

    PHYSICAL REVIEW E　95　(1)　012131　2017/01

    DOI： 10.1103/PhysRevE.95.012131 　

    ISSN： 2470-0045

    eISSN： 2470-0053

  19. Motif analysis for small-number effects in chemical reaction dynamics Peer-reviewed

    Nen Saito, Yuki Sughiyama, Kunihiko Kaneko

    JOURNAL OF CHEMICAL PHYSICS　145　(9)　094111　2016/09

    DOI： 10.1063/1.4961675 　

    ISSN： 0021-9606

    eISSN： 1089-7690

  20. Discreteness-induced transitions in multibody reaction systems Peer-reviewed

    Yohei Saito, Yuki Sughiyama, Kunihiko Kaneko, Tetsuya J. Kobayashi

    PHYSICAL REVIEW E　94　(2)　022140　2016/08

    DOI： 10.1103/PhysRevE.94.022140 　

    ISSN： 2470-0045

    eISSN： 2470-0053

  21. Relaxation of the EM Algorithm via Quantum Annealing for Gaussian Mixture Models Peer-reviewed

    Hideyuki Miyahara, Koji Tsumura, Yuki Sughiyama

    2016 IEEE 55TH CONFERENCE ON DECISION AND CONTROL (CDC)　4674-4679　2016

    ISSN： 0743-1546

  22. 細胞間相互作用による細胞集団の自己制御に関する数理的解析

    三木 翔太, 斎藤 陽平, 杉山 友規, 小林 徹也

    生研研究　68　(3)　245-246　2016

  23. Deterministic quantum annealing EM algorithm for discrete latent variables Peer-reviewed

    Hideyuki Miyahara, Koji Tsumura, Yuki Sughiyama

    2016 IEEE 55th Conference on Decision and Control (CDC)　4674-4679　2016

  24. Fluctuation Relations of Fitness and Information in Population Dynamics Peer-reviewed

    Tetsuya J. Kobayashi, Yuki Sughiyama

    PHYSICAL REVIEW LETTERS　115　(23)　238102　2015/12

    DOI： 10.1103/PhysRevLett.115.238102 　

    ISSN： 0031-9007

    eISSN： 1079-7114

  25. Pathwise thermodynamic structure in population dynamics Peer-reviewed

    Yuki Sughiyama, Tetsuya J. Kobayashi, Koji Tsumura, Kazuyuki Aihara

    Physical Review E - Statistical, Nonlinear, and Soft Matter Physics　91　(3)　2015/03/12

    Publisher: American Physical Society

    DOI： 10.1103/PhysRevE.91.032120 　

    ISSN： 1550-2376 1539-3755

  26. Pathwise thermodynamic structure in population dynamics Peer-reviewed

    Yuki Sughiyama, Tetsuya J. Kobayashi, Koji Tsumura, Kazuyuki Aihara

    PHYSICAL REVIEW E　91　(3)　032120　2015/03

    DOI： 10.1103/PhysRevE.91.032120 　

    ISSN： 2470-0045

    eISSN： 2470-0053

  27. 18aCX-10 Clausius inequality in population dynamics

    Sughiyama Y., Kobayashi T. J.

    Meeting Abstracts of the Physical Society of Japan　70　2780-2780　2015

    Publisher: The Physical Society of Japan (JPS)

    DOI： 10.11316/jpsgaiyo.70.2.0_2780 　

    ISSN： 2189-079X

  28. 2P273 Fitness Value of Information in Biological Systems(24. Mathematical biology,Poster)

    Kobayashi Tetsuya J., Sughiyama Yuki

    Seibutsu Butsuri　54　(1)　S240　2014

    Publisher: The Biophysical Society of Japan General Incorporated Association

    DOI： 10.2142/biophys.54.S240_3 　

  29. 2P279 Mathematical Analysis of Small Number Effect in Biochemical Reactions(24. Mathematical biology,Poster)

    Saito Nen, Sughiyama Yuki, Kaneko Kunihiko

    Seibutsu Butsuri　54　(1)　S241　2014

    Publisher: The Biophysical Society of Japan General Incorporated Association

    DOI： 10.2142/biophys.54.S241_3 　

  30. Variational Principle in Langevin Processes (<Special Issue>The 4th Young Scientist Meeting on Statistical Physics and Information Processing in Sendai)

    Ohzeki Masayuki, Sughiyama Yuki

    Interdisciplinary information sciences　19　93-99　2013/08

    Publisher: Tohoku University

    DOI： 10.4036/iis.2013.93 　

    ISSN： 1340-9050

    More details Close

    The recent work, Nemoto and Sasa [Phys. Rev. E, 83: 030105(R) (2011)], has shown that large deviations of the current characterizing a nonequilibrium system are obtained by observing the typical current for a modified system specified by a variational principle. In the present study, we will give a generalized version of the Nemoto–Sasa study by extracting a hidden mathematical structure from the fluctuation-response relation which is well-known in statistical mechanics. Here, the minimization of the Kullback–Leibler divergence plays an essential role.

  31. Nonequilibrium work relation in a macroscopic system Peer-reviewed

    Yuki Sughiyama, Masayuki Ohzeki

    JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT　P04012　2013/04

    DOI： 10.1088/1742-5468/2013/04/P04012 　

    ISSN： 1742-5468

  32. Variational principle in Langevin processes Peer-reviewed

    Yuki Sughiyama, Masayuki Ohzeki

    Interdisciplinary Information Sciences　19　93　2013

  33. Extended Jarzynski equality in general Langevin system Peer-reviewed

    Yuki Sughiyama, Masayuki Ohzeki

    PHYSICA E-LOW-DIMENSIONAL SYSTEMS & NANOSTRUCTURES　43　(3)　790-793　2011/01

    DOI： 10.1016/j.physe.2010.07.053 　

    ISSN： 1386-9477

  34. Fluctuation theorem for the renormalized entropy change in the strongly nonlinear nonequilibrium regime Peer-reviewed

    Yuki Sughiyama, Sumiyoshi Abe

    PHYSICAL REVIEW E　78　(2)　021101　2008/08

    DOI： 10.1103/PhysRevE.78.021101 　

    ISSN： 1539-3755

  35. Macroscopic proof of the Jarzynski-Wojcik fluctuation theorem for heat exchange Peer-reviewed

    Yuki Sughiyama, Sumiyoshi Abe

    JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT　P05008　2008/05

    DOI： 10.1088/1742-5468/2008/05/P05008 　

    ISSN： 1742-5468

  36. 揺らぎ定理（熱場の量子論とその応用、研究会報告）

    杉山 友規

    素粒子論研究　116　(2)　B102-B109　2008

Show all ︎Show first 5

Misc. 2

  1. Generative modeling from an optimal control perspective

    清水怜央, 大関真之, 大関真之, 大関真之, 杉山友規

    日本物理学会講演概要集(CD-ROM)　79　(2)　2024

    ISSN： 2189-079X

  2. Deterministic quantum annealing expectation-maximization algorithm (vol 11, 113404, 2017)

    Hideyuki Miyahara, Koji Tsumura, Yuki Sughiyama

    JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT　2020　(10)　2020/10

    DOI： 10.1088/1742-5468/abb017 　

    ISSN： 1742-5468

Presentations 26

  1. 経路積分的見方による年齢構造付き個体群動態の解析とその応用

    杉山 友規, 中島 蒼, 小林 徹也

    数理生物学会（JSMB2019）　2019/09

  2. A large deviation theory for an age-structured population dynamics and its application to inference of cell states Invited

    2019/09

  3. Pathwise analysis for a structured population dynamics and its application to inference of cell states

    The XXVII International Conference on Statistical Physics (Statphys27)　2019/07

  4. Pathwise analysis for a multi-type age-structured population dynamics International-presentation Invited

    Yuki Sughiyama

    Eurasian Health and Medicine 2018　2018/11

  5. Retrospective Approach for Age Structured Population Dynamics International-presentation Invited

    杉山 友規

    The 12th AIMS Conference on Dynamical Systems, Differential Equations and Applications　2018/07

  6. 時間遡及的見方による集団増殖率の解析 Invited

    杉山 友規

    第70回日本人口学会研究大会　2018/06

  7. Semi-Markov過程上の大偏差関数とその増殖系への応用

    杉山 友規, 小林 徹也

    日本物理学会 2018年年次大会　2018/03

  8. A retrospective analysis of the multi-state age-structured population dynamics International-presentation Invited

    杉山 友規

    Ancestral lines in populations under selection　2017/11

  9. Path-wise analysis for the age-structured population dynamics International-presentation

    杉山 友規

    Large Deviation Theory in Statistical Physics: Recent Advances and Future Challenges　2017/10

  10. age構造付き増殖過程の解析

    杉山 友規, 小林 徹也

    日本物理学会 2017年年次大会　2017/03

  11. Age-structured population dynamics with type switching International-presentation Invited

    杉山 友規

    Frontier in Quantitative Understanding of Dynamic Living States & its Applications　2017/03

  12. Steady State Thermodynamic Structure in Population Dynamics International-presentation Invited

    杉山 友規

    Physical approaches for growing & evolving populations　2017/02

  13. Thermodynamic Structure in Population Dynamics International-presentation Invited

    杉山 友規

    Japan q-bio week, WS: entropy, information and control　2016/01

  14. 集団増殖系におけるClausius不等式

    杉山 友規, 小林 徹也

    日本物理学会 2015年秋季大会　2015/09

  15. 集団増殖率における変分構造とその応用 Invited

    杉山 友規, 小林 徹也, 津村 幸治, 合原 一幸

    第2回制御部門マルチシンポジウム（MSCS2015）　2015/03

  16. 増殖過程における変分構造

    杉山 友規, 小林 徹也, 津村 幸治, 合原 一幸

    日本物理学会 2014年秋季大会　2014/09

  17. Thermodynamic Structure in Population Dynamics Invited

    杉山 友規

    Multidisciplinary Approach Forum（M A F 2 0 1 4）　2014/09

  18. 集団増殖率に見る熱力学構造 Invited

    杉山 友規, 小林 徹也, 津村 幸治, 合原 一幸

    第１回制御部門マルチシンポジウム（MSCS2014）　2014/03

  19. 大偏差を用いた高次応答の解析

    杉山 友規

    日本物理学会 2013年年次大会　2013/03

  20. Nemoto-Sasa理論の幾何学的解釈

    杉山 友規

    4th YSM-SPIP in Sendai/Prologue Series IV of FSPIP　2012/12

  21. 揺らぎ定理に基づく熱力学の構成 Invited

    杉山 友規

    基研研究会「情報統計力学の最前線」（YSM-SPIP2012）　2012/03

  22. 準安定状態を持つ系に対する仕事の関係式

    杉山 友規, 大関 真之

    日本物理学会 2011年秋季大会　2011/09

  23. Ito過程に従う多体模型の熱力学極限 Invited

    杉山 友規

    YSM-SPIP2011　2011/03

  24. 確率過程論から見たJarzynski 恒等式 Invited

    杉山 友規

    YSM-SPIP2010　2010/03

  25. 揺らぎ定理

    杉山 友規

    第18回統計物理学研究会　2007/10

  26. 揺らぎ定理について Invited

    杉山 友規

    基研研究会「熱場の量子論とその応用」　2007/09

Show all Show first 5

Research Projects 2

  1. 集団増殖系に内在する定常状態熱力学構造とその応用 Competitive

    杉山 友規

    Offer Organization: 日本学術振興会

    System: 科学研究費助成事業（若手研究Ｂ）

    2016/04 - 2019/03

  2. 非平衡系における熱力学理論の構築とその応用 Competitive

    杉山 友規

    Offer Organization: 日本学術振興会

    System: 科学研究費補助金（特別研究員奨励費）

    2010/04 - 2012/03

Teaching Experience 3

  1. 解析学基礎　東京工科大学コンピュータサイエンス学部

  2. 線形代数I　東京工科大学コンピュータサイエンス学部

  3. 線形代数Ⅱ　東京工科大学コンピュータサイエンス学部