复杂系统中的因果涌现研究综述

王志鹏; 张江

doi:10.12202/j.0476-0301.2023049

复杂系统中的因果涌现研究综述

doi: 10.12202/j.0476-0301.2023049

王志鹏¹,
张江^{1, 2, ,}

1.
北京师范大学系统科学学院，100875，北京
2.
北京集智研究院，100085，北京

基金项目: 国家自然科学基金资助项目（61673070）

详细信息

通讯作者:
张江（1978—），博士，教授，博士生导师. 研究方向：复杂系统建模、因果涌现、规模法则、机器学习. E-mail：zhangjiang@bnu.edu.cn

中图分类号: N949
计量
- 文章访问数: 79
- HTML全文浏览量: 45
- PDF下载量: 28
- 被引次数: 0
出版历程
- 收稿日期: 2023-04-04
- 录用日期: 2023-09-06
- 网络出版日期: 2023-09-21
- 刊出日期: 2023-10-31

Causal emergence in complex systems

WANG Zhipeng¹,
ZHANG Jiang^{1, 2
, ,}

1.
School of Systems Science，Beijing Normal University，100875，Beijing，China
2.
Swarma Research，100085，Beijing，China

摘要

摘要: 聚焦复杂系统：就涌现、因果以及因果涌现进行了定量描述；阐述了因果与涌现的联系；系统梳理了基于粗粒化和信息分解2种定量刻画因果涌现的方法，以及基于信息分解和神经信息压缩2种因果涌现辨识方法；详细介绍了各方法的基本原理、优缺点，以及相关应用等．基于粗粒化方法定义了因果涌现框架，其可应用于离散动力系统；利用信息分解方法求解时结果会依赖冗余信息；基于信息分解所提出的指标可识别数据中的因果涌现，且找到一个充分条件；基于数据驱动的神经信息压缩方法可扩展并应用于连续动力系统，并可自动提取不同层级的粗粒化函数，以及构建不同层级的动力学，还可识别不同类型动力学系统的因果涌现；改进现有方法并用于识别更为复杂的系统与学习自动化分组，解决通用动力学大模型等问题．
- 复杂系统 /
- 因果涌现 /
- 有效信息 /
- 粗粒化 /
- 信息分解 /
- 神经信息压缩
Abstract: Focusing on complex systems: a quantitative description is provided for emergence, causality, and causal emergence．The relationship between causality and emergence is elucidated． Two quantitative methods, coarse-graining and information decomposition, are reviewed for characterizing causal emergence． Two causal emergence identification methods based on information decomposition and neural information compression are presented． Basic principles, advantages, disadvantages, and relevant applications of each method are detailed． The framework of causal emergence is defined based on coarse-graining method, which can be applied to discrete dynamical systems． The use of information decomposition method in solving problems relies on redundant information． The proposed indicators based on information decomposition can identify causal emergence in data and find a sufficient condition． The data-driven neural information compression method is scalable and applicable to continuous dynamical systems． It can automatically extract coarse-grained functions at different levels and construct hierarchical dynamics, and identify causal emergence in different types of dynamical systems． Existing methods are improved and applied to identify more complex systems, automate learning groups, and address challenges in general dynamics of large-scale models and other related problems．
- complex system /
- causal emergence /
- effective information /
- coarse-graining /
- information decomposition /
- neural information compression

HTML全文

图 1 抽象的因果涌现框架

下载: 全尺寸图片幻灯片

图 2 神经信息压缩方法的模型框架示意

下载: 全尺寸图片幻灯片

参考文献(52)

[1]	智韬,司光亚,贺筱媛. 电力关键基础设施网络仿真模型研究[J]. 系统仿真学报,2010,22(11):2732 doi: 10.16182/j.cnki.joss.2010.11.019
[2]	叶青. 基于复杂网络理论的轨道交通网络脆弱性分析[J]. 中国安全科学学报,2012,22(2):122 doi: 10.3969/j.issn.1003-3033.2012.02.020
[3]	KWAPIEŃ J,DROŻDŻ S. Physical approach to complex systems[J]. Physics Reports,2012,515(3/4):115
[4]	王飞跃,史帝夫·兰森. 从人工生命到人工社会:复杂社会系统研究的现状和展望[J]. 复杂系统与复杂性科学,2004,1(1):33 doi: 10.3969/j.issn.1672-3813.2004.01.007
[5]	王飞跃. 人工社会、计算实验、平行系统:关于复杂社会经济系统计算研究的讨论[J]. 复杂系统与复杂性科学,2004,1(4):25 doi: 10.3969/j.issn.1672-3813.2004.04.002
[6]	彭雁虹,褚启勤,李怀祖. 生命系统理论及其应用综述[J]. 系统工程理论与实践,1997,17(3):35
[7]	ANDERSON P W. More is different[J]. Science,1972,177(4047):393 doi: 10.1126/science.177.4047.393
[8]	CAMAZINE S,DENEUBOURG J L,FRANKS N R,et al. Self-organization in biological systems[M]. New Jersey:Princeton University Press,2020
[9]	苗东升. 论系统思维(六):重在把握系统的整体涌现性[J]. 系统科学学报,2006,14(1):1
[10]	方锦清,汪小帆,刘曾荣. 略论复杂性问题和非线性复杂网络系统的研究[J]. 科技导报,2004,22(2):9 doi: 10.3321/j.issn:1000-7857.2004.02.003
[11]	MEEHL P E,SELLARS W. The concept of emergence[J]. Minnesota Studies in the Philosophy of Science,1956,1:239
[12]	BEDAU M A. Weak emergence[J]. Philosophical Perspectives,1997,11:375
[13]	PEPPER S C. Emergence[J]. The Journal of Philosophy,1926,23(9):241 doi: 10.2307/2014779
[14]	FROMM J. Types and forms of emergence[EB/OL]. (2005-06-13)[2023-02-15]. https://arxiv.org/abs/nlin/0506028.pdf
[15]	PEARL J. Graphs,causality,and structural equation models[J]. Sociological Methods & Research,1998,27(2):226
[16]	GRANGER C W J. Investigating causal relations by econometric models and cross-spectral methods[J]. Econometrica,1969,37(3):424 doi: 10.2307/1912791
[17]	NEUBERG L G. Causality:models,reasoning,and inference,by Judea Pearl,Cambridge University Press,2000[J]. Econometric Theory,2003,19:675
[18]	HOEL E P,ALBANTAKIS L,TONONI G. Quantifying causal emergence shows that macro can beat micro[J]. Proceedings of the National Academy of Sciences of the United States of America,2013,110(49):19790
[19]	HOEL E. When the map is better than the territory[J]. Entropy,2017,19(5):188 doi: 10.3390/e19050188
[20]	ROSAS F E,MEDIANO P A M,JENSEN H J,et al. Reconciling emergences:an information-theoretic approach to identify causal emergence in multivariate data[J]. PLoS Computational Biology,2020,16(12):e1008289 doi: 10.1371/journal.pcbi.1008289
[21]	ZHANG J A,LIU K W. Neural information squeezer for causal emergence[J]. Entropy,2022,25(1):26 doi: 10.3390/e25010026
[22]	MARSH H W,YEUNG A S. Top-down,bottom-up,and horizontal models:the direction of causality in multidimensional,hierarchical self-concept models[J]. Journal of Personality and Social Psychology,1998,75(2):509 doi: 10.1037/0022-3514.75.2.509
[23]	CHVYKOV P,HOEL E. Causal geometry[J]. Entropy,2020,23(1):24 doi: 10.3390/e23010024
[24]	VARLEY T,HOEL E. Emergence as the conversion of information:a unifying theory[EB/OL]. (2021-04-27)[2023-02-16].https://arxiv.org/abs/2104.13368.pdf
[25]	KLEIN B,HOEL E. The emergence of informative higher scales in complex networks[J]. Complexity,2020,2020:1
[26]	GRIEBENOW R,KLEIN B,HOEL E. Finding the right scale of a network:efficient identification of causal emergence through spectral clustering[EB/OL]. (2022-02-16)[2023-02-17]. https://arxiv.org/abs/1908.07565v1
[27]	汪小帆,李翔,陈关荣. 复杂网络理论及其应用[M]. 北京:清华大学出版社,2006
[28]	MARROW S,MICHAUD E J,HOEL E. Examining the causal structures of deep neural networks using information theory[J]. Entropy,2020,22(12):1429 doi: 10.3390/e22121429
[29]	KLEIN B,HOEL E,SWAIN A,et al. Evolution and emergence:higher order information structure in protein interactomes across the tree of life[J]. Integrative Biology,2021,13(12):283 doi: 10.1093/intbio/zyab020
[30]	BRENNAN K,ANSHUMAN S,TRAVIS B,et al. Exploring noise,degeneracy and determinism in biological networks with the einet package[J]. Methods in Ecology and Evolution,2022,13(4):799 doi: 10.1111/2041-210X.13805
[31]	MEDIANO P A M,ROSAS F E,LUPPI A I,et al. Greater than the parts:a review of the information decomposition approach to causal emergence [EB/OL]. (2022-05-23)[2023-02-17]. https://royalsocietypublishing.org/doi/abs/10.1098/rsta.2021.0246
[32]	TONONI G,SPORNS O. Measuring information integration[J]. BMC Neuroscience,2003,4:31 doi: 10.1186/1471-2202-4-31
[33]	WILLIAMS P L,BEER R D. Nonnegative decomposition of multivariate information[EB/OL]. (2010-04-14)[2023-02-18].https://arxiv.org/abs/1004.2515.pdf
[34]	ROSAS F E,MEDIANO P A M,RASSOULI B,et al. An operational information decomposition via synergistic disclosure[J]. Journal of Physics A:Mathematical and Theoretical,2020,53(48):485001 doi: 10.1088/1751-8121/abb723
[35]	REICHSTEIN M,CAMPS-VALLS G,STEVENS B,et al. Deep learning and process understanding for data-driven Earth system science[J]. Nature,2019,566(7743):195 doi: 10.1038/s41586-019-0912-1
[36]	SENIOR A W,EVANS R,JUMPER J,et al. Improved protein structure prediction using potentials from deep learning[J]. Nature,2020,577(7792):706 doi: 10.1038/s41586-019-1923-7
[37]	KEISER M J,SETOLA V,IRWIN J J,et al. Predicting new molecular targets for known drugs[J]. Nature,2009,462(7270):175 doi: 10.1038/nature08506
[38]	程乐峰,余涛,张孝顺,等. 机器学习在能源与电力系统领域的应用和展望[J]. 电力系统自动化,2019,43(1):15
[39]	张伟伟,朱林阳,刘溢浪,等. 机器学习在湍流模型构建中的应用进展[J]. 空气动力学学报,2019,37(3):444
[40]	GLYMOUR C,ZHANG K,SPIRTES P. Review of causal discovery methods based on graphical models[EB/OL]. (2019-06-04)[2023-02-18]. https://royalsocietypublishing.org/doi/abs/10.1098/rsta.2021.0246
[41]	CASADIEGO J,NITZAN M,HALLERBERG S,et al. Model-free inference of direct network interactions from nonlinear collective dynamics[J]. Nature Communications,2017,8:2192 doi: 10.1038/s41467-017-02288-4
[42]	ZHANG Y,GUO Y,ZHANG Z,et al. Universal framework for reconstructing complex networks and node dynamics from discrete or continuous dynamics data[EB/OL]. (2022-09-16)[2023-02-18]. https://journals.aps.org/pre/abstract/10.1103/PhysRevE.106.034315
[43]	GLYMOUR C,ZHANG K,SPIRTES P. Review of causal discovery methods based on graphical models[J]. Frontiers in Genetics,2019,10:524 doi: 10.3389/fgene.2019.00524
[44]	蔡瑞初,陈薇,张坤,等. 基于非时序观察数据的因果关系发现综述[J]. 计算机学报,2017,40(6):1470 doi: 10.11897/SP.J.1016.2017.01470
[45]	HU H Y,LI S H,WANG L,et al. Machine learning holographic mapping by neural network renormalization group[J]. Physical Review Research,2020,2(2):023369 doi: 10.1103/PhysRevResearch.2.023369
[46]	LI S H,WANG L. Neural network renormalization group[J]. Physical Review Letters,2018,121(26):260601 doi: 10.1103/PhysRevLett.121.260601
[47]	DINH L,SOHL-DICKSTEIN J,BENGIO S. Density estimation using Real NVP[EB/OL]. (2017-02-27)[2023-02-18].https://arxiv.org/abs/1605.08803
[48]	HOEL E,LEVIN M. Emergence of informative higher scales in biological systems:a computational toolkit for optimal prediction and control[J]. Communicative & Integrative Biology,2020,13(1):108
[49]	SWAIN A,WILLIAMS S D,DI FELICE L J,et al. Interactions and information:exploring task allocation in ant colonies using network analysis[J]. Animal Behaviour,2022,189:69 doi: 10.1016/j.anbehav.2022.04.015
[50]	SCHÖLKOPF B,LOCATELLO F,BAUER S,et al. Toward causal representation learning[J]. Proceedings of the IEEE,2021,109(5):612 doi: 10.1109/JPROC.2021.3058954
[51]	IWASAKI Y,SIMON H A. Causality and model abstraction[J]. Artificial Intelligence,1994,67(1):143 doi: 10.1016/0004-3702(94)90014-0
[52]	MATSUO Y,LECUN Y,SAHANI M,et al. Deep learning,reinforcement learning,and world models[J]. Neural Networks,2022,152:267 doi: 10.1016/j.neunet.2022.03.037