Oracle bounds and exact algorithm for dyadic classification trees

被引:12
作者
Blanchard, G
Schäfer, C
Rozenholc, Y
机构
[1] FIRST, Fraunhofer Inst, D-12489 Berlin, Germany
[2] Univ Paris 06, Lab Probabil & Modeles Aleatoires, F-75252 Paris 05, France
来源
LEARNING THEORY, PROCEEDINGS | 2004年 / 3120卷
关键词
D O I
10.1007/978-3-540-27819-1_26
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
This paper introduces a new method using dyadic decision trees for estimating a classification or a regression function in a multi-class classification problem. The estimator is based on model selection by penalized empirical loss minimization. Our work consists in two complementary parts: first, a theoretical analysis of the method leads to deriving oracle-type inequalities for three different possible loss functions. Secondly, we present an algorithm able to compute the estimator in an exact way.
引用
收藏
页码:378 / 392
页数:15
相关论文
共 16 条
[1]  
ADELSONVELSKII GM, 1962, SOV MATH DOKL, V3, P1259
[2]   Risk bounds for model selection via penalization [J].
Barron, A ;
Birgé, L ;
Massart, P .
PROBABILITY THEORY AND RELATED FIELDS, 1999, 113 (03) :301-413
[3]   APPROXIMATION OF DENSITY-FUNCTIONS BY SEQUENCES OF EXPONENTIAL-FAMILIES [J].
BARRON, AR ;
SHEU, CH .
ANNALS OF STATISTICS, 1991, 19 (03) :1347-1369
[4]  
BLANCHARD G, 2004, STAT PERFORMANCE SUP
[5]  
Breiman L., 1998, CLASSIFICATION REGRE
[6]   Histograms selection with an Akaike type criterion [J].
Castellan, G .
COMPTES RENDUS DE L ACADEMIE DES SCIENCES SERIE I-MATHEMATIQUE, 2000, 330 (08) :729-732
[7]  
Devroye L., 1996, APPL MATH, V31
[8]   Cart and best-ortho-basis: A connection [J].
Donoho, DL .
ANNALS OF STATISTICS, 1997, 25 (05) :1870-1911
[9]  
Gey S, 2003, LECT NOTES STAT, V171, P369
[10]  
KLEMELA J, 2003, MULTIVARIATE HISTOGR