• • 上一篇    

竞争性因果学习及其应用

李爱忠1, 任若恩2, 董纪昌3   

  1. 1. 山西财经大学财政与公共经济学院, 太原 030006;
    2. 北京航空航天大学经济管理学院, 北京 100191;
    3. 中国科学院大学经济与管理学院, 北京 100190
  • 收稿日期:2021-09-08 修回日期:2022-04-24 发布日期:2022-12-13
  • 基金资助:
    国家社会科学基金(19BTJ026)资助课题.

李爱忠, 任若恩, 董纪昌. 竞争性因果学习及其应用[J]. 系统科学与数学, 2022, 42(11): 3015-3026.

LI Aizhong, REN Ruoen, DONG Jichang. Competitive Causal Learning and Its Application[J]. Journal of Systems Science and Mathematical Sciences, 2022, 42(11): 3015-3026.

Competitive Causal Learning and Its Application

LI Aizhong1, REN Ruoen2, DONG Jichang3   

  1. 1. School of Public Finance & Economics, Shanxi University of Finance and Economics, Taiyuan 030006;
    2. School of Economics and Management, Beihang University, Beijing 100191;
    3. School of Economics and Management, University of Chinese Academy of Sciences, Beijing 100190
  • Received:2021-09-08 Revised:2022-04-24 Published:2022-12-13
因果关系的研究一直紧密围绕人类探索世界和发现世 界的主题, 传统的仅仅研究事物之间的统计相关关系作用有限, 很难满 足经济社会快速发展的需要. 文章将自监督学习和对抗学习结合, 利用 图网络模型和系统动力学的反演模型, 从大规模无监督数据中挖掘潜在 的隐含信息, 基于对比约束, 构建物理驱动与数据驱动的统一框架, 然 后采用极大极小博弈策略学习不同因果模态的一致性表征, 从而逼近真正 的因果关系, 为揭示潜藏在数据背后的内在规律提供了有力的分析工具. 文章 将非随机因果学习思想融入机器学习框架, 对克服现有深度学习在抽象、推 理及神经网络可解释性等方面的不足具有重要指导意义.
The study of causality has always been the theme of human exploration of the world. The traditional study of only the correlation between things can no longer meet the needs of current economic and social development. This paper combines self-supervised learning and adversarial learning, and uses Bayesian network model and system dynamics inversion model to mine its own supervised information from large-scale unsupervised data. Based on comparative constraints, we build a unified framework of physical drive and data drive, and then use minimax game strategies to learn consistent representations of different causal modes. Therefore, the true causality can be approximated. It provides a powerful analysis tool for revealing the inherent laws hidden behind the data. This paper integrates the idea of non-random causal learning into the machine learning framework, which has important guiding significance for overcoming the shortcomings of existing deep learning in abstraction, reasoning and neural network interpretability.

MR(2010)主题分类: 

()
[1] Lauritzen S L. Discussion on causality. Blackwell Publishing Ltd, 2004, 31: 189-193.
[2] 苗旺, 刘春辰, 耿直. 因果推断的统计方法. 中国科学: 数学, 2018, 48: 1753-1778. (Miao W, Liu C C, Geng Z. Statistical approaches for causal inference. Sci. Sin. Math, 2018, 48: 1753-1778.)
[3] Angrist J D, Imbens G W. Identification and estimation of local average treatment effects. Econometrica, 1994, 62(2): 467-475.
[4] Spirtes P. Introduction to causal inference. J. Mach. Learn. Res., 2010, 11: 1643-1662.
[5] Bartlett M S. Methods of estimating mental factors. Nature, 1938, 141: 609-610.
[6] Hochreiter S, Clevert D A, Obermayer K. A new summarization method for Affymetrix probe level data. Bioinformatics, 2006, 22: 943-949
[7] Hirzel A H, Hausser J, Chessel D, et al. Ecological-niche factor analysis: How to compute habitat-suitability maps without absence data? Ecology, 2002, 83: 2027-2036.
[8] Bai J, Li K. Statistical analysis of factor models of high dimension. Annals of Statistics, 2012, 40: 436-465.
[9] 李爱忠, 任若恩, 李泽楷, 等. 高维稀疏低秩的多目标矩阵回归模型及其组合管理策略. 系统工 程理论与实践, 2020, 40(9): 2292-2301. (Li A Z, Ren R E, Li Z K, et al. High-dimensional sparse low-rank multi-objective matrix regression model and its portfolio management strategy. Systems Engineering — Theory & Practice, 2020, 40(9): 2292-2301.)
[10] Athey S, Eckles D, Imbens G W. Exact $p$-values for network interference. J Amer. Statist. Assoc., 2018, 113(521): 230-240.
[11] Sun B, Liu L, Miao W, et al. Semiparametric estimation with data missing not at random using an instrumental variable. Statist Sinica, 2018, 28(4): 1965-1983.
[12] Sofer T, Richardson D B, Colicino E, et al. On negative outcome control of unobserved confounding as a generalization of difference-in-differences. Statist Sci., 2016, 31(3): 348-361.
[13] Miao W, Geng Z, Tchetgen TE. Identifying causal effects with proxy variables of an unmeasured confounder. Biometrika, 2018, 105(4): 987-993.
[14] Pearl J. Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference. San Francisco: Morgan Kaufmann, 1988.
[15] Pearl J. Causality: Models, Reasoning, and Inference. 2nd Ed. New York: Cambridge University Press, 2009.
[16] Hyttinen A, Hoyer P O, Eberhardt F, et al. Discovering cyclic causal models with latent variables: A general SAT-based procedure. Proceedings of the 29th Conference on Uncertainty in Articial Intelligence. Corvallis: AUAI Press, 2013, 301-310.
[17] Borboudakis G, Tsamardinos I. Towards robust and versatile causal discovery for business applications. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: Amer. Math Soc., 2016, 1435-1444.
[18] Fei N, Yang Y. Estimating linear causality in the presence of latent variables. Cluster Computing, 2017, 20(2): 1025-1033.
[19] 张朝阳, 陈阳, 弭元元,等. 从数据到结构——动力学网络重构. 中国科学: 物理 力学 天文学, 2020, 50: 010502. (Zhang Z Y, Chen Y, Mi Y Y, et al. From data to network structure — Reconstruction of dynamic networks. Sci. Sin. Phys. Mech. Astron., 2020, 50: 010502.)
[20] 李爱忠, 任若恩, 李泽楷, 等.图嵌入下稀疏低秩集成预测的多因子资产选择策略.中 国管理科学, 2021, 29(11): 23-32. (Li A Z, Ren R E, Li Z K, et al. Multi-factor portfolio selection strategy based on sparse low-rank ensemble prediction under graph embedding. Chinese Journal of Management Science, 2021, 29(11): 23-32.
[1] 朱宏泉, 陈诗韵, 李亚静. 腐败对企业股权融资成本的影响研究——基于新股发行的视角[J]. 系统科学与数学, 2022, 42(9): 2412-2431.
[2] 何志强, 陈凤娥, 彭幸春. 均值-方差准则下保险公司带内幕信息的时间一致投资-再保险策略[J]. 系统科学与数学, 2022, 42(9): 2432-2447.
[3] 吴胤昊, 陈荣达, 汪圣楠, 俞静婧. 随机利率随机波动率混合指数跳扩散模型下的期权定价[J]. 系统科学与数学, 2022, 42(8): 2207-2234.
[4] 王宗润, 谭郭玺. 基于确定性等价进行等权重调整的组合投资策略研究[J]. 系统科学与数学, 2022, 42(2): 287-303.
[5] 梁进, 周汇慧. 资产波动率为随机的信用等级迁移风险评估模型[J]. 系统科学与数学, 2022, 42(2): 304-317.
[6] 陈晓红,彭宛露,田美玉. 基于投资者情绪的股票价格及成交量预测研究[J]. 系统科学与数学, 2016, 36(12): 2294-2306.
[7] 冯玲,吴江樵. 股票市场组合的市场风险度量研究------基于相关性模型[J]. 系统科学与数学, 2016, 36(12): 2307-2324.
[8] 江良,林鸿熙,宋丽平. 基于P-样条方法的短期利率模型非参估计[J]. 系统科学与数学, 2016, 36(11): 2137-2150.
[9] 刘凌晨,段白鸽. 一类随机利率下特殊年金的计算[J]. 系统科学与数学, 2016, 36(10): 1670-1677.
[10] 宋丽平,余王辉. 经理期权整体实施与非限制实施的等价[J]. 系统科学与数学, 2016, 36(10): 1710-1720.
[11] 龚玉婷,郑旭. 基于Copula模型的尾部相依性长记忆效应研究[J]. 系统科学与数学, 2016, 36(6): 783-799.
[12] 唐振鹏,黄友柏,罗雪玲. 资产组合优化的多分形模型及实证分析[J]. 系统科学与数学, 2016, 36(2): 198-.
[13] 张涤新,眭以宁. 深圳股票市场的日内流动性研究[J]. 系统科学与数学, 2015, 35(12): 1487-1500.
[14] 丁元耀,卢祖帝. 安全首要准则下的借贷约束与不确定风险投资[J]. 系统科学与数学, 2015, 35(12): 1529-1545.
[15] 唐勇,黄志刚. 多分形视角下的金融市场波动建模研究[J]. 系统科学与数学, 2015, 35(6): 667-684.
阅读次数
全文


摘要