Fast computation of cross-validated properties in full linear leave-many-out procedures

Emili Besalú

doi:10.1023/A:1010924406885

Fast computation of cross-validated properties in full linear leave-many-out procedures

Emili Besalú

Institute of Computational Chemistry and Catalysis

Research output: Contribution to journal › Scientific article › peer-review

43 Citations (Scopus)

Abstract

A general theorem which allows the fast and direct computation of predicted properties in a full multiple linear leave-many-out procedure is demonstrated by induction. The result allows the description of a general algorithm which only requires a single multiple linear regression calculation. From the data generated by this fitting, in a full leave-n-out procedure involving a set of m objects, the resolution of (^m_n) linear systems of equations of dimension n × n suffices to obtain all the sets of cross-validated properties.

Original language	English
Pages (from-to)	191-204
Number of pages	14
Journal	Journal of Mathematical Chemistry
Volume	29
Issue number	3
DOIs	https://doi.org/10.1023/A:1010924406885
Publication status	Published - Apr 2001

Keywords

Cross-validation
Leave-many-out
Leave-n-out theorem
Leave-one-out
Multiple linear regression

Access to Document

10.1023/A:1010924406885

Cite this

@article{6ef80931356144b59c2e5b3f37d63b86,

title = "Fast computation of cross-validated properties in full linear leave-many-out procedures",

abstract = "A general theorem which allows the fast and direct computation of predicted properties in a full multiple linear leave-many-out procedure is demonstrated by induction. The result allows the description of a general algorithm which only requires a single multiple linear regression calculation. From the data generated by this fitting, in a full leave-n-out procedure involving a set of m objects, the resolution of (mn) linear systems of equations of dimension n × n suffices to obtain all the sets of cross-validated properties.",

keywords = "Cross-validation, Leave-many-out, Leave-n-out theorem, Leave-one-out, Multiple linear regression",

author = "Emili Besal{\'u}",

year = "2001",

month = apr,

doi = "10.1023/A:1010924406885",

language = "English",

volume = "29",

pages = "191--204",

journal = "Journal of Mathematical Chemistry",

issn = "0259-9791",

publisher = "Springer Netherlands",

number = "3",

}

TY - JOUR

T1 - Fast computation of cross-validated properties in full linear leave-many-out procedures

AU - Besalú, Emili

PY - 2001/4

Y1 - 2001/4

N2 - A general theorem which allows the fast and direct computation of predicted properties in a full multiple linear leave-many-out procedure is demonstrated by induction. The result allows the description of a general algorithm which only requires a single multiple linear regression calculation. From the data generated by this fitting, in a full leave-n-out procedure involving a set of m objects, the resolution of (mn) linear systems of equations of dimension n × n suffices to obtain all the sets of cross-validated properties.

AB - A general theorem which allows the fast and direct computation of predicted properties in a full multiple linear leave-many-out procedure is demonstrated by induction. The result allows the description of a general algorithm which only requires a single multiple linear regression calculation. From the data generated by this fitting, in a full leave-n-out procedure involving a set of m objects, the resolution of (mn) linear systems of equations of dimension n × n suffices to obtain all the sets of cross-validated properties.

KW - Cross-validation

KW - Leave-many-out

KW - Leave-n-out theorem

KW - Leave-one-out

KW - Multiple linear regression

UR - https://www.scopus.com/pages/publications/0035536399

U2 - 10.1023/A:1010924406885

DO - 10.1023/A:1010924406885

M3 - Scientific article

AN - SCOPUS:0035536399

SN - 0259-9791

VL - 29

SP - 191

EP - 204

JO - Journal of Mathematical Chemistry

JF - Journal of Mathematical Chemistry

IS - 3

ER -

Fast computation of cross-validated properties in full linear leave-many-out procedures

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this