Banzhaf random forests: Cooperative game theory based random forests with consistency

被引:45
作者
Sun, Jianyuan [1 ]
Zhong, Guoqiang [1 ]
Huang, Kaizhu [2 ]
Dong, Junyu [1 ]
机构
[1] Ocean Univ China, Dept Comp Sci & Technol, 238 Songling Rd, Qingdao 266100, Peoples R China
[2] Xian Jiaotong Liverpool Univ, Dept Elect & Elect Engn, SIP, Suzhou 215123, Peoples R China
基金
国家重点研发计划; 中国国家自然科学基金; 对外科技合作项目(国际科技项目);
关键词
Random forests; Cooperative game; Banzhaf power index; Consistency; DECISION FORESTS; NEURAL-NETWORK; CLASSIFICATION; REGRESSION; RECOGNITION; FRAMEWORK; ENSEMBLES;
D O I
10.1016/j.neunet.2018.06.006
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Random forests algorithms have been widely used in many classification and regression applications. However, the theory of random forests lags far behind their applications. In this paper, we propose a novel random forests classification algorithm based on cooperative game theory. The Banzhaf power index is employed to evaluate the power of each feature by traversing possible feature coalitions. Hence, we call the proposed algorithm Banzhaf random forests (BRFs). Unlike the previously used information gain ratio, which only measures the power of each feature for classification and pays less attention to the intrinsic structure of the feature variables, the Banzhaf power index can measure the importance of each feature by computing the dependency among the group of features. More importantly, we have proved the consistency of BRFs, which narrows the gap between the theory and applications of random forests. Extensive experiments on several UCI benchmark data sets and three real world applications show that BRFs perform significantly better than existing consistent random forests on classification accuracy, and better than or at least comparable with Breiman's random forests, support vector machines (SVMs) and k-nearest neighbors (KNNs) classifiers. (C) 2018 Elsevier Ltd. All rights reserved.
引用
收藏
页码:20 / 29
页数:10
相关论文
共 48 条
[1]   Shape quantization and recognition with randomized trees [J].
Amit, Y ;
Geman, D .
NEURAL COMPUTATION, 1997, 9 (07) :1545-1588
[2]   An ensemble of dynamic neural network identifiers for fault detection and isolation of gas turbine engines [J].
Amozegar, M. ;
Khorasani, K. .
NEURAL NETWORKS, 2016, 76 :106-121
[3]  
[Anonymous], PROBABILISTIC THEORY
[4]  
[Anonymous], SPRINGER SCI BUSINES
[5]  
[Anonymous], INT C MACH LEARN
[6]  
[Anonymous], 2013, P 30 INT C MACH LEAR
[7]  
[Anonymous], 2011, ACM T INTEL SYST TEC, DOI DOI 10.1145/1961189.1961199
[8]  
[Anonymous], IEEE C COMP VIS PATT
[9]   Hybrid extreme rotation forest [J].
Ayerdi, Borja ;
Grana, Manuel .
NEURAL NETWORKS, 2014, 52 :33-42
[10]  
Banzhaf III J. F., 1964, RUTGERS LAW REV, V19, P317