Recursive approach for constructing the q = 1/2 maximum entropy distribution from redundant data

L. Rebollo-Neira*, A. Plastino

*Corresponding author for this work

Research output: Contribution to journalArticle

Abstract

A recursive approach for computing the q = 1/2 nonextensive maximum entropy distribution of the previously introduced formalism for data subset selection is proposed. Such an approach is based on an iterative biorthogonalization technique, which allows for the incorporation of the Lagrange multipliers that determine the distribution to the workings of the algorithm devised for selecting relevant data subsets. This technique circumvents the necessity of inverting operators and yields a recursive procedure to appropriately modify the Lagrange multipliers so as to account for each new constraint.

Original languageEnglish
Article number032102
JournalPhysical Review E
Volume66
Issue number3
DOIs
Publication statusPublished - 1 Sep 2002

Fingerprint

Lagrange multipliers
Maximum Entropy
set theory
Nonextensive Entropy
entropy
Subset Selection
formalism
operators
Subset
Computing
Operator
Necessity

Bibliographical note

©2002 American Physical Society. Recursive approach for constructing the qÄ1Õ2 maximum entropy distribution
from redundant data
L. Rebollo-Neira and A. Plastino
Phys. Rev. E 66, 032102 – Published 27 September 2002

Cite this

@article{dd011703e89d45ff835d3199f1beb6dc,
title = "Recursive approach for constructing the q = 1/2 maximum entropy distribution from redundant data",
abstract = "A recursive approach for computing the q = 1/2 nonextensive maximum entropy distribution of the previously introduced formalism for data subset selection is proposed. Such an approach is based on an iterative biorthogonalization technique, which allows for the incorporation of the Lagrange multipliers that determine the distribution to the workings of the algorithm devised for selecting relevant data subsets. This technique circumvents the necessity of inverting operators and yields a recursive procedure to appropriately modify the Lagrange multipliers so as to account for each new constraint.",
author = "L. Rebollo-Neira and A. Plastino",
note = "{\circledC}2002 American Physical Society. Recursive approach for constructing the q{\"A}1{\~O}2 maximum entropy distribution from redundant data L. Rebollo-Neira and A. Plastino Phys. Rev. E 66, 032102 – Published 27 September 2002",
year = "2002",
month = "9",
day = "1",
doi = "10.1103/PhysRevE.66.032102",
language = "English",
volume = "66",
journal = "Physical Review E",
issn = "1539-3755",
publisher = "American Physical Society",
number = "3",

}

Recursive approach for constructing the q = 1/2 maximum entropy distribution from redundant data. / Rebollo-Neira, L.; Plastino, A.

In: Physical Review E, Vol. 66, No. 3, 032102, 01.09.2002.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Recursive approach for constructing the q = 1/2 maximum entropy distribution from redundant data

AU - Rebollo-Neira, L.

AU - Plastino, A.

N1 - ©2002 American Physical Society. Recursive approach for constructing the qÄ1Õ2 maximum entropy distribution from redundant data L. Rebollo-Neira and A. Plastino Phys. Rev. E 66, 032102 – Published 27 September 2002

PY - 2002/9/1

Y1 - 2002/9/1

N2 - A recursive approach for computing the q = 1/2 nonextensive maximum entropy distribution of the previously introduced formalism for data subset selection is proposed. Such an approach is based on an iterative biorthogonalization technique, which allows for the incorporation of the Lagrange multipliers that determine the distribution to the workings of the algorithm devised for selecting relevant data subsets. This technique circumvents the necessity of inverting operators and yields a recursive procedure to appropriately modify the Lagrange multipliers so as to account for each new constraint.

AB - A recursive approach for computing the q = 1/2 nonextensive maximum entropy distribution of the previously introduced formalism for data subset selection is proposed. Such an approach is based on an iterative biorthogonalization technique, which allows for the incorporation of the Lagrange multipliers that determine the distribution to the workings of the algorithm devised for selecting relevant data subsets. This technique circumvents the necessity of inverting operators and yields a recursive procedure to appropriately modify the Lagrange multipliers so as to account for each new constraint.

UR - http://www.scopus.com/inward/record.url?scp=37649030678&partnerID=8YFLogxK

UR - https://journals.aps.org/pre/abstract/10.1103/PhysRevE.66.032102

U2 - 10.1103/PhysRevE.66.032102

DO - 10.1103/PhysRevE.66.032102

M3 - Article

C2 - 12366163

AN - SCOPUS:37649030678

VL - 66

JO - Physical Review E

JF - Physical Review E

SN - 1539-3755

IS - 3

M1 - 032102

ER -