Fixed-size kernel models with SVD truncation schemes

Ricardo Castro; Siamak Mehrkanoon; Anna Marconato; Joannes Schoukens; Johan Suykens

Fixed-size kernel models with SVD truncation schemes

Ricardo Castro, Siamak Mehrkanoon, Anna Marconato, Joannes Schoukens, Johan Suykens

Elektriciteit

Onderzoeksoutput: Meeting abstract (Book)

Samenvatting

Criteria for assessing the generalization performance take the form Prediction Error = training error + complexity term. The complexity term represents a penalty growing with the number of free parameters in the model. We consider different versions of ﬁxed-size kernel models related to ﬁxed-size least squares support vector machines (FS-LSSVM) and consider the effective degrees of freedom (EDF) as the notion for model complexity.

Originele taal-2	English
Titel	Benelux meeting, Heijden, The Netherlands, March 25-27, 2014
Pagina's	77
Aantal pagina's	1
Status	Published - 25 mrt. 2014
Evenement	33rd Benelux Meeting on Systems and Control - Heijen, Netherlands Duur: 25 mrt. 2014 → 27 mrt. 2014

Conference

Conference	33rd Benelux Meeting on Systems and Control
Land/Regio	Netherlands
Stad	Heijen
Periode	25/03/14 → 27/03/14

Citeer dit

@inbook{aee3f53667134872b499b106133d1e68,

title = "Fixed-size kernel models with SVD truncation schemes",

abstract = "Criteria for assessing the generalization performance take the form Prediction Error = training error + complexity term. The complexity term represents a penalty growing with the number of free parameters in the model. We consider different versions of ﬁxed-size kernel models related to ﬁxed-size least squares support vector machines (FS-LSSVM) and consider the effective degrees of freedom (EDF) as the notion for model complexity.",

keywords = "Prediction error, effective degrees of freedem, model complexity",

author = "Ricardo Castro and Siamak Mehrkanoon and Anna Marconato and Joannes Schoukens and Johan Suykens",

year = "2014",

month = mar,

day = "25",

language = "English",

isbn = "978-90-365-3597-7",

pages = "77",

booktitle = "Benelux meeting, Heijden, The Netherlands, March 25-27, 2014",

note = "33rd Benelux Meeting on Systems and Control ; Conference date: 25-03-2014 Through 27-03-2014",

}

TY - CHAP

T1 - Fixed-size kernel models with SVD truncation schemes

AU - Castro, Ricardo

AU - Mehrkanoon, Siamak

AU - Marconato, Anna

AU - Schoukens, Joannes

AU - Suykens, Johan

PY - 2014/3/25

Y1 - 2014/3/25

N2 - Criteria for assessing the generalization performance take the form Prediction Error = training error + complexity term. The complexity term represents a penalty growing with the number of free parameters in the model. We consider different versions of ﬁxed-size kernel models related to ﬁxed-size least squares support vector machines (FS-LSSVM) and consider the effective degrees of freedom (EDF) as the notion for model complexity.

AB - Criteria for assessing the generalization performance take the form Prediction Error = training error + complexity term. The complexity term represents a penalty growing with the number of free parameters in the model. We consider different versions of ﬁxed-size kernel models related to ﬁxed-size least squares support vector machines (FS-LSSVM) and consider the effective degrees of freedom (EDF) as the notion for model complexity.

KW - Prediction error

KW - effective degrees of freedem

KW - model complexity

M3 - Meeting abstract (Book)

SN - 978-90-365-3597-7

SP - 77

BT - Benelux meeting, Heijden, The Netherlands, March 25-27, 2014

T2 - 33rd Benelux Meeting on Systems and Control

Y2 - 25 March 2014 through 27 March 2014

ER -

Fixed-size kernel models with SVD truncation schemes

Samenvatting

Conference

Vingerafdruk

Citeer dit