Dynamic Random Forests

被引：86

作者：

Bernard, Simon ^{[1
]}

Adam, Sebastien ^{[1
]}

Heutte, Laurent ^{[1
]}

机构：

[1] Univ Rouen, LITIS EA 4108, F-76801 St Etienne, France

来源：

PATTERN RECOGNITION LETTERS | 2012年 / 33卷 / 12期

关键词：

Random forests; Ensemble of classifiers; Random feature selection; Dynamic induction;

D O I：

10.1016/j.patrec.2012.04.003

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper, we introduce a new Random Forest (RF) induction algorithm called Dynamic Random Forest (DRF) which is based on an adaptative tree induction procedure. The main idea is to guide the tree induction so that each tree will complement as much as possible the existing trees in the ensemble. This is done here through a resampling of the training data, inspired by boosting algorithms, and combined with other randomization processes used in traditional RF methods. The DRF algorithm shows a significant improvement in terms of accuracy compared to the standard static RF induction algorithm. (C) 2012 Elsevier B.V. All rights reserved.

引用

页码：1580 / 1586

页数：7

共 36 条

[1] Shape quantization and recognition with randomized trees [J].

Amit, Y ;

Geman, D .

NEURAL COMPUTATION, 1997, 9 (07) :1545-1588

[2]

Banfield RE, 2004, LECT NOTES COMPUT SC, V3077, P223

[3] A comparison of decision tree ensemble creation techniques [J].

Banfield, Robert E. ;

Hall, Lawrence O. ;

Bowyer, Kevin W. ;

Kegelmeyer, W. P. .

IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2007, 29 (01) :173-180

[4]

BERNARD S, 2009, THESIS U ROUEN FRANC

[5]

Bernard S, 2008, LECT NOTES COMPUT SC, V5227, P430, DOI 10.1007/978-3-540-85984-0_52

[6]

Bernard S, 2007, PROC INT CONF DOC, P1043

[7]

Bernard S, 2009, LECT NOTES COMPUT SC, V5519, P171, DOI 10.1007/978-3-642-02326-2_18

[8] Random forests [J].

Breiman, L .

MACHINE LEARNING, 2001, 45 (01) :5-32

[9] Random forests [J].

Breiman, L .

MACHINE LEARNING, 2001, 45 (01) :5-32

[10]

Breiman L, 1998, ANN STAT, V26, P801

← 1 2 3 4 →