On the generalization ability of GRLVQ networks

被引：44

作者：

Hammer, B ^{[1
]}

Strickert, M

Villmann, T

机构：

[1] Univ Osnabruck, Dept Math Comp Sci, LNM, D-4500 Osnabruck, Germany

[2] Univ Leipzig, Clin Psychotherapy, Leipzig, Germany

来源：

NEURAL PROCESSING LETTERS | 2005年 / 21卷 / 02期

关键词：

adaptive metric; generalization bounds; LVQ; margin optimization; relevance LVQ;

D O I：

10.1007/s11063-004-1547-1

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We derive a generalization bound for prototype-based classifiers with adaptive metric. The bound depends on the margin of the classifier and is independent of the dimensionality of the data. It holds for classifiers based on the Euclidean metric extended by adaptive relevance terms. In particular, the result holds for relevance learning vector quantization (RLVQ) [4] and generalized relevance learning vector quantization (GRLVQ) [19].

引用

页码：109 / 120

页数：12

共 40 条

[1] ANGUITA D, 2002, ICANN 2002, P1345
[2] [Anonymous], 1994, MACHINE LEARNING NEU
[3] [Anonymous], SELF ORGANIZING MAPS
[4] [Anonymous], Pattern Recognition With Fuzzy Objective Function Algorithms
[5] Bartlett P. L., 2003, Journal of Machine Learning Research, V3, P463, DOI 10.1162/153244303321897690
[6] BOJER T, 2001, P EUR S ART NEUR NET, P271
[7] BOUSQUET O, IN PRESS ADV NEURAL
[8] Comparison of adaptive methods for function estimation from samples
Cherkassky, V
Gehring, D
[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS, 1996, 7 (04): : 969 - 984
[9] CORTES C, 1995, MACH LEARN, V20, P273, DOI 10.1023/A:1022627411411
[10] CRAMMER K, IN PRESS ADV NEURAL

← 1 2 3 4 →