Classification with non-i.i.d. sampling

Zheng-Chu Guo; Lei Shi

doi:10.1016/j.mcm.2011.03.042

Classification with non-i.i.d. sampling

Zheng-Chu Guo
, Lei Shi

Department of Mathematics

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

15 Citations (Scopus)

Abstract

We study learning algorithms for classification generated by regularization schemes in reproducing kernel Hilbert spaces associated with a general convex loss function in a non-i.i.d. process. Error analysis is studied and our main purpose is to provide an elaborate capacity dependent error bounds by applying concentration techniques involving the ℓ²-empirical covering numbers. © 2011 Elsevier Ltd.

Original language	English
Pages (from-to)	1347-1364
Journal	Mathematical and Computer Modelling
Volume	54
Issue number	5-6
DOIs	https://doi.org/10.1016/j.mcm.2011.03.042
Publication status	Published - Sept 2011

Bibliographical note

Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

Research Keywords

Capacity dependent error bounds
Learning theory
Regularized classification
Reproducing kernel Hilbert spaces
β-mixing sequence

Access Output

10.1016/j.mcm.2011.03.042Licence: Elsevier user license

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

@article{b0d3d97a890c42559dfae53c35028e4c,

title = "Classification with non-i.i.d. sampling",

abstract = "We study learning algorithms for classification generated by regularization schemes in reproducing kernel Hilbert spaces associated with a general convex loss function in a non-i.i.d. process. Error analysis is studied and our main purpose is to provide an elaborate capacity dependent error bounds by applying concentration techniques involving the ℓ2-empirical covering numbers. {\textcopyright} 2011 Elsevier Ltd.",

keywords = "Capacity dependent error bounds, Learning theory, Regularized classification, Reproducing kernel Hilbert spaces, β-mixing sequence",

author = "Zheng-Chu Guo and Lei Shi",

note = "Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].",

year = "2011",

month = sep,

doi = "10.1016/j.mcm.2011.03.042",

language = "English",

volume = "54",

pages = "1347--1364",

journal = "Mathematical and Computer Modelling",

issn = "0895-7177",

publisher = "Elsevier Ltd.",

number = "5-6",

}

TY - JOUR

T1 - Classification with non-i.i.d. sampling

AU - Guo, Zheng-Chu

AU - Shi, Lei

N1 - Publication details (e.g. title, author(s), publication statuses and dates) are captured on an “AS IS” and “AS AVAILABLE” basis at the time of record harvesting from the data source. Suggestions for further amendments or supplementary information can be sent to [email protected].

PY - 2011/9

Y1 - 2011/9

N2 - We study learning algorithms for classification generated by regularization schemes in reproducing kernel Hilbert spaces associated with a general convex loss function in a non-i.i.d. process. Error analysis is studied and our main purpose is to provide an elaborate capacity dependent error bounds by applying concentration techniques involving the ℓ2-empirical covering numbers. © 2011 Elsevier Ltd.

AB - We study learning algorithms for classification generated by regularization schemes in reproducing kernel Hilbert spaces associated with a general convex loss function in a non-i.i.d. process. Error analysis is studied and our main purpose is to provide an elaborate capacity dependent error bounds by applying concentration techniques involving the ℓ2-empirical covering numbers. © 2011 Elsevier Ltd.

KW - Capacity dependent error bounds

KW - Learning theory

KW - Regularized classification

KW - Reproducing kernel Hilbert spaces

KW - β-mixing sequence

UR - https://www.scopus.com/pages/publications/79957878487

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-79957878487&origin=recordpage

U2 - 10.1016/j.mcm.2011.03.042

DO - 10.1016/j.mcm.2011.03.042

M3 - RGC 21 - Publication in refereed journal

SN - 0895-7177

VL - 54

SP - 1347

EP - 1364

JO - Mathematical and Computer Modelling

JF - Mathematical and Computer Modelling

IS - 5-6

ER -

Classification with non-i.i.d. sampling

Abstract

Bibliographical note

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this