− | [[复杂网络中的因果涌现]]的识别困难在于系统性、自动搜索所有潜在的粗粒化策略<ref>Varley, T.; Hoel, E. Emergence as the conversion of information: A unifying theory. arXiv 2021, arXiv:2104.13368.</ref><ref>Chvykov, P.; Hoel, E. Causal Geometry. Entropy 2021, 23, 24. [CrossRef]</ref><ref name=":1">Rosas, F.E.; Mediano, P.A.M.; Jensen, H.J.; Seth, A.K.; Barrett, A.B.; Carhart-Harris, R.L.; Bor, D. Reconciling emergences: An information-theoretic approach to identify causal emergence in multivariate data. PLoS Comput. Biol. 2020, 16, e1008289. [CrossRef]</ref><ref>Varley, T.F. Flickering emergences: The question of locality in information-theoretic approaches to emergence. arXiv 2022, arXiv:2208.14502.</ref>。Klein的方法通过节点聚类<ref name=":2">Klein, B.; Swain, A.; Byrum, T.; Scarpino, S.V.; Fagan, W.F. Exploring noise, degeneracy and determinism in biological networks with the einet package. Methods Ecol. Evol. 2022, 13, 799–804.</ref><ref>Klein, B.; Hoel, E. The Emergence of Informative Higher Scales in Complex Networks. Complexity 2020, 2020, 8932526. [CrossRef]</ref>提升EI,但假设底层节点动力学是扩散的,未考虑真实系统中更复杂的动力学。即使节点分组已知,粗粒化策略仍需考虑簇中所有节点的微观与宏观状态映射。 | + | [[复杂网络中的因果涌现]]的识别困难在于系统性、自动搜索所有潜在的粗粒化策略<ref>Varley, T.; Hoel, E. Emergence as the conversion of information: A unifying theory. arXiv 2021, arXiv:2104.13368.</ref><ref>Chvykov, P.; Hoel, E. Causal Geometry. Entropy 2021, 23, 24. [CrossRef]</ref><ref name=":1">Rosas, F.E.; Mediano, P.A.M.; Jensen, H.J.; Seth, A.K.; Barrett, A.B.; Carhart-Harris, R.L.; Bor, D. Reconciling emergences: An information-theoretic approach to identify causal emergence in multivariate data. PLoS Comput. Biol. 2020, 16, e1008289. [CrossRef]</ref><ref>Varley, T.F. Flickering emergences: The question of locality in information-theoretic approaches to emergence. arXiv 2022, arXiv:2208.14502.</ref>。Klein的方法通过节点聚类<ref name=":2" /><ref>Klein, B.; Hoel, E. The Emergence of Informative Higher Scales in Complex Networks. Complexity 2020, 2020, 8932526. [CrossRef]</ref>提升EI,但假设底层节点动力学是扩散的,未考虑真实系统中更复杂的动力学。即使节点分组已知,粗粒化策略仍需考虑簇中所有节点的微观与宏观状态映射。 |
| 近年来,机器学习得到长足发展,其跨学科应用也逐渐出现<ref>Silver, D.; Schrittwieser, J.; Simonyan, K.; Antonoglou, I.; Huang, A.; Guez, A.; Hubert, T.; Baker, L.; Lai, M.; Bolton, A.; et al. Mastering the game of Go without human knowledge. Nature 2017, 550, 354–359.</ref><ref>LeCun,Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444.</ref><ref>Reichstein, M.; Camps-Valls, G.; Stevens, B.; Jung, M.; Denzler, J.; Carvalhais, N. Deep learning and process understanding for data-driven Earth system science. Nature 2019, 566, 195–204.</ref><ref>Senior, A.W.; Evans, R.; Jumper, J.; Kirkpatrick, J.; Sifre, L.; Green, T.; Qin, C.; Žídek, A.; Nelson, A.W.R.; Bridgland, A.; et al. Improved protein structure prediction using potentials from deep learning. Nature 2020, 577, 706–710.</ref>。由此方法,以数据为驱动的、自动发现因果涌现<ref>Tank, A.; Covert, I.; Foti, N.; Shojaie, A.; Fox, E. Neural Granger Causality. arXiv 2018, arXiv:1802.05842.</ref><ref>Löwe,S.; Madras, D.; Zemel, R.; Welling, M. Amortized causal discovery: Learning to infer causal graphs from time-series data. arXiv 2020, arXiv:2006.10833.</ref><ref>Glymour, C.; Zhang, K.; Spirtes, P. Review of Causal Discovery Methods Based on Graphical Models. Front. Genet. 2019, 10, 524.</ref><ref>Casadiego, J.; Nitzan, M.; Hallerberg, S.; Timme, M. Model-free inference of direct network interactions from nonlinear collective dynamics. Nat. Commun. 2017, 8, 1–10.</ref>,甚至复杂系统的动力学已成为可能<ref>Sanchez-Gonzalez, A.; Heess, N.; Springenberg, J.T.; Merel, J.; Riedmiller, M.; Hadsell, R.; Battaglia, P. Graph networks as learnable physics engines for inference and control. In Proceedings of the International Conference on Machine Learning, Stockholm, Sweden, 10–15 July 2018 ; pp. 4470–4479.</ref><ref>Zhang, Z.; Zhao, Y.; Liu, J.; Wang, S.; Tao, R.; Xin, R.; Zhang, J. A general deep learning framework for network reconstruction and dynamics learning. Appl. Netw. Sci. 2019, 4, 1–17.</ref><ref>Kipf, T.; Fetaya, E.; Wang, K.C.; Welling, M.; Zemel, R. Neural relational inference for interacting systems. In Proceedings of the International Conference on Machine Learning, Stockholm, Sweden, 10–15 July 2018; pp. 2688–2697.</ref><ref>Chen,B.; Huang, K.; Raghupathi, S.; Chandratreya, I.; Du, Q.; Lipson, H. Discovering State Variables Hidden in Experimental Data. arXiv 2021, arXiv:2112.10755.</ref>。因果涌现识别问题可表述为“在微观动力学精确预测的约束下,最大化宏观动力学的有效信息(EI)”。神经信息压缩器(NIS)是解决此问题的通用机器学习框架。NIS通过可逆神经网络建模粗粒化策略<ref>Koch-Janusz, M.; Ringel, Z. Mutual information, neural networks and the renormalization group. Nat. Phys. 2018, 14, 578–582.</ref><ref name=":3">Li, S.H.; Wang, L. Neural Network Renormalization Group. Phys. Rev. Lett. 2018, 121, 260601.</ref><ref>Hu,H.Y.; Li, S.H.; Wang, L.; You, Y.Z. Machine learning holographic mapping by neural network renormalization group. Phys. Rev. Res. 2020, 2, 023369.</ref><ref name=":4">Hu,H.; Wu,D.; You, Y.Z.; Olshausen, B.; Chen, Y. RG-Flow: A hierarchical and explainable flow model based on renormalization group and sparse prior. Mach. Learn. Sci. Technol. 2022, 3, 035009.</ref><ref>Gökmen,D.E.; Ringel, Z.; Huber, S.D.; Koch-Janusz, M. Statistical physics through the lens of real-space mutual information. Phys. Rev. Lett. 2021, 127, 240603.</ref>,将任意<math>\mathcal{R}^p</math>到<math>\mathcal{R}^q(q \leq p)</math> 的映射分解为一系列信息转换和弃用过程,可对整个框架进行数学分析。 | | 近年来,机器学习得到长足发展,其跨学科应用也逐渐出现<ref>Silver, D.; Schrittwieser, J.; Simonyan, K.; Antonoglou, I.; Huang, A.; Guez, A.; Hubert, T.; Baker, L.; Lai, M.; Bolton, A.; et al. Mastering the game of Go without human knowledge. Nature 2017, 550, 354–359.</ref><ref>LeCun,Y.; Bengio, Y.; Hinton, G. Deep learning. Nature 2015, 521, 436–444.</ref><ref>Reichstein, M.; Camps-Valls, G.; Stevens, B.; Jung, M.; Denzler, J.; Carvalhais, N. Deep learning and process understanding for data-driven Earth system science. Nature 2019, 566, 195–204.</ref><ref>Senior, A.W.; Evans, R.; Jumper, J.; Kirkpatrick, J.; Sifre, L.; Green, T.; Qin, C.; Žídek, A.; Nelson, A.W.R.; Bridgland, A.; et al. Improved protein structure prediction using potentials from deep learning. Nature 2020, 577, 706–710.</ref>。由此方法,以数据为驱动的、自动发现因果涌现<ref>Tank, A.; Covert, I.; Foti, N.; Shojaie, A.; Fox, E. Neural Granger Causality. arXiv 2018, arXiv:1802.05842.</ref><ref>Löwe,S.; Madras, D.; Zemel, R.; Welling, M. Amortized causal discovery: Learning to infer causal graphs from time-series data. arXiv 2020, arXiv:2006.10833.</ref><ref>Glymour, C.; Zhang, K.; Spirtes, P. Review of Causal Discovery Methods Based on Graphical Models. Front. Genet. 2019, 10, 524.</ref><ref>Casadiego, J.; Nitzan, M.; Hallerberg, S.; Timme, M. Model-free inference of direct network interactions from nonlinear collective dynamics. Nat. Commun. 2017, 8, 1–10.</ref>,甚至复杂系统的动力学已成为可能<ref>Sanchez-Gonzalez, A.; Heess, N.; Springenberg, J.T.; Merel, J.; Riedmiller, M.; Hadsell, R.; Battaglia, P. Graph networks as learnable physics engines for inference and control. In Proceedings of the International Conference on Machine Learning, Stockholm, Sweden, 10–15 July 2018 ; pp. 4470–4479.</ref><ref>Zhang, Z.; Zhao, Y.; Liu, J.; Wang, S.; Tao, R.; Xin, R.; Zhang, J. A general deep learning framework for network reconstruction and dynamics learning. Appl. Netw. Sci. 2019, 4, 1–17.</ref><ref>Kipf, T.; Fetaya, E.; Wang, K.C.; Welling, M.; Zemel, R. Neural relational inference for interacting systems. In Proceedings of the International Conference on Machine Learning, Stockholm, Sweden, 10–15 July 2018; pp. 2688–2697.</ref><ref>Chen,B.; Huang, K.; Raghupathi, S.; Chandratreya, I.; Du, Q.; Lipson, H. Discovering State Variables Hidden in Experimental Data. arXiv 2021, arXiv:2112.10755.</ref>。因果涌现识别问题可表述为“在微观动力学精确预测的约束下,最大化宏观动力学的有效信息(EI)”。神经信息压缩器(NIS)是解决此问题的通用机器学习框架。NIS通过可逆神经网络建模粗粒化策略<ref>Koch-Janusz, M.; Ringel, Z. Mutual information, neural networks and the renormalization group. Nat. Phys. 2018, 14, 578–582.</ref><ref name=":3">Li, S.H.; Wang, L. Neural Network Renormalization Group. Phys. Rev. Lett. 2018, 121, 260601.</ref><ref>Hu,H.Y.; Li, S.H.; Wang, L.; You, Y.Z. Machine learning holographic mapping by neural network renormalization group. Phys. Rev. Res. 2020, 2, 023369.</ref><ref name=":4">Hu,H.; Wu,D.; You, Y.Z.; Olshausen, B.; Chen, Y. RG-Flow: A hierarchical and explainable flow model based on renormalization group and sparse prior. Mach. Learn. Sci. Technol. 2022, 3, 035009.</ref><ref>Gökmen,D.E.; Ringel, Z.; Huber, S.D.; Koch-Janusz, M. Statistical physics through the lens of real-space mutual information. Phys. Rev. Lett. 2021, 127, 240603.</ref>,将任意<math>\mathcal{R}^p</math>到<math>\mathcal{R}^q(q \leq p)</math> 的映射分解为一系列信息转换和弃用过程,可对整个框架进行数学分析。 |
| 因果表征学习旨在提取观测数据背后的因果隐变量<ref>Chalupka, K.; Eberhardt, F.; Perona, P. Causal feature learning: An overview. Behaviormetrika 2017, 44, 137–164.</ref><ref>Schölkopf, B.; Locatello, F.; Bauer, S.; Ke, N.R.; Kalchbrenner, N.; Goyal, A.; Bengio, Y. Toward causal representation learning. Proc. IEEE 2021, 109, 612–634.</ref>,编码过程可理解为粗粒化。因果涌现识别与因果表征学习相似,但目标不同:前者寻找更优粗粒化策略,后者提取数据中的因果关系。多尺度建模和粗粒化操作引入了新的理论问题<ref>Iwasaki, Y.; Simon, H.A. Causality and model abstraction. Artif. Intell. 1994, 67, 143–194.</ref><ref>Rubenstein, P.K.; Weichwald, S.; Bongers, S.; Mooij, J.; Janzing, D.; Grosse-Wentrup, M.; Schölkopf, B. Causal consistency of structural equation models. arXiv 2017, arXiv:1707.00819.</ref><ref>Beckers, S.; Eberhardt, F.; Halpern, J.Y. Approximate causal abstractions. In Proceedings of the Uncertainty in Artificial Intelligence, Virtual, 3–6 August 2020; pp. 606–615.</ref>。 | | 因果表征学习旨在提取观测数据背后的因果隐变量<ref>Chalupka, K.; Eberhardt, F.; Perona, P. Causal feature learning: An overview. Behaviormetrika 2017, 44, 137–164.</ref><ref>Schölkopf, B.; Locatello, F.; Bauer, S.; Ke, N.R.; Kalchbrenner, N.; Goyal, A.; Bengio, Y. Toward causal representation learning. Proc. IEEE 2021, 109, 612–634.</ref>,编码过程可理解为粗粒化。因果涌现识别与因果表征学习相似,但目标不同:前者寻找更优粗粒化策略,后者提取数据中的因果关系。多尺度建模和粗粒化操作引入了新的理论问题<ref>Iwasaki, Y.; Simon, H.A. Causality and model abstraction. Artif. Intell. 1994, 67, 143–194.</ref><ref>Rubenstein, P.K.; Weichwald, S.; Bongers, S.; Mooij, J.; Janzing, D.; Grosse-Wentrup, M.; Schölkopf, B. Causal consistency of structural equation models. arXiv 2017, arXiv:1707.00819.</ref><ref>Beckers, S.; Eberhardt, F.; Halpern, J.Y. Approximate causal abstractions. In Proceedings of the Uncertainty in Artificial Intelligence, Virtual, 3–6 August 2020; pp. 606–615.</ref>。 |