|Table of Contents|

Fast causal network skeleton learning algorithm


Research Field:
Publishing date:


Fast causal network skeleton learning algorithm
Hong Yinghan
School of Physics and Electronic Engineering,Hanshan Normal University,Chaozhou 521041,China
causal networks skeletons high dimensional networks causal sets conditional independence tests
Aiming at the problem that the traditional structure of causal network learning algorithms is unfit for high dimensional networks,a fast causal network skeleton construction algorithm is proposed for high dimensional networks.A mutual information accelerating strategy is used based on the maximum dependence and the minimum redundancy,two candidate causal sets between two nodes are found out,and conditional independence tests of the 2 nodes are proposed in the union of the 2 candidate causal sets.Real data experiments show that the time complexity of the algorithm proposed here is better than the traditional algorithms for high dimensional networks; the recognition accuracy of networks is improved because conditional independence tests are decreased.


[1] Pearl J.Causality:Models,reasoning and inference[M].Cambridge,USA:The MIT Press,2009.
[2]Spirtes,Glymour C,Scheines R.Causation,prediction,and search[M].Cambridge,USA:The MIT Press,2000.
[3]Tsamardinos I,Brown L E,Aliferis C F.The max-min hill-climbing Bayesian network structure learning algorithm[J].Machine Learning,2006,65(1):31-78.
[4]Ma Saisai,Li Jiuyong,Liu Lin,et al.Mining combined causes in large data sets[J].Knowledge-Based Systems,2016,92:104-111.
[5]Chickering D M.Optimal structure identification with greedy search[J].Journal of Machine Learning Research,2002,3(3):507-554.
[6]Vargas P,Moioli R,De Castro L N,et al.Artificial homeostatic system:A novel approach[C]//Proceedings of Advances in Artifical Life.Berlin,Germany:Springer Berlin Heidelberg,2005:754-764.
[7]Shimizu S,Hoyer P O,Hyvarinen A,et al.A linear non-Gaussian acyclic model for causal discovery[J].Journal of Machine Learning Research,2006,7(4):2003-2030.
Zhang Hao,Hao Zhifeng,Cai Ruichu,et al.High dimensional causality discovering based on mutual information[J].Application Research of Computers,2015,32(2):382-385.
[9]Hoyer P O,Janzing D,Mooij J M,et al.Nonlinear causal discovery with additive noise models[C]//Advances in Neural Information Processing Systems.Vancouver,Canada:MIT Press,2009:689-696.
[10]Peters J,Janzing D,Schoelkopf B.Causal inference on discrete data using additive noise models[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2011,33(12):2436-2450.
Zhang Hao,Hao Zhifeng,Cai Ruichu,et al.An approach for inferring causal directions in high dimensional causal networks[J].Journal of Chinese Computer Systems,2015,36(6):1358-1362.
[12]Janzing D,Mooij J,Zhang Kun,et al.Information-geometric approach to inferring causal directions[J].Artificial Intelligence,2012,182(10):1-31.
[13]Peng H,Long F,Ding C.Variable selection based on mutual information criteria of max-dependency,max-relevance,and min-redundancy[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,27(8):1226-1238.


Last Update: 2016-06-30