LEARNING RATES FOR DENSITY LEVEL DETECTION

被引：6

作者：

Scovel, Clint ^{[1
]}

Hush, Don ^{[1
]}

Steinwart, Ingo ^{[1
]}

机构：

[1] Los Alamos Natl Lab, CCS 3, Modeling Algorithms & Informat Grp, Los Alamos, NM 87545 USA

来源：

ANALYSIS AND APPLICATIONS | 2005年 / 3卷 / 04期

关键词：

Anomaly detection; learning theory; rates; density level detection;

D O I：

10.1142/S0219530505000625

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

In this paper, we address learning rates for the density level detection (DLD) problem. We begin by proving a "No Free Lunch Theorem" showing that rates cannot be obtained in general. Then, we apply a recently established classification framework to obtain rates for DLD support vector machines under mild assumptions on the density.

引用

页码：357 / 371

页数：15

共 14 条

[1] Learning distributions by their density levels: A paradigm for learning without a teacher [J].

BenDavid, S ;

Lindenbaum, M .

JOURNAL OF COMPUTER AND SYSTEM SCIENCES, 1997, 55 (01) :171-182

[2] Cluster analysis: a further approach based on density estimation [J].

Cuevas, A ;

Febrero, M ;

Fraiman, R .

COMPUTATIONAL STATISTICS & DATA ANALYSIS, 2001, 36 (04) :441-459

[3]

Devroye L., 1996, A probabilistic theory of pattern recognition

[4]

Hartigan J. A., 1987, J AM STAT ASSOC, V82, P267, DOI DOI 10.1080/01621459.1987.10478428

[5]

Hartigan J. A., 1975, CLUSTERING ALGORITHM

[6] EXCESS MASS ESTIMATES AND TESTS FOR MULTIMODALITY [J].

MULLER, DW ;

SAWITZKI, G .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1991, 86 (415) :738-746

[7] MEASURING MASS CONCENTRATIONS AND ESTIMATING DENSITY CONTOUR CLUSTERS - AN EXCESS MASS APPROACH [J].

POLONIK, W .

ANNALS OF STATISTICS, 1995, 23 (03) :855-881

[8]

SAWITZKI G, 1996, DATA KNOWLEDGE THEOR, P203

[9]

Steinwart I, 2005, J MACH LEARN RES, V6, P211

[10]

STEINWART I, 2005, NEURAL INF PROCESS S, V17, P1337

← 1 2 →