Reparametrization-covariant theory for on-line learning of probability distributions

Research output: Contribution to journalArticle

Abstract

We discuss the on-line learning of probability distributions in a reparametrization covariant formulation. Reparametrization covariance plays an essential role not only to respect an intrinsic property of “information” but also for pattern recognition problems. We can obtain an optimal on-line learning algorithm with reparametrization invariance, where the conformal gauge connects a covariant formulation with a noncovariant one in a natural way.

Original languageEnglish
Number of pages1
JournalPhysical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics
Volume64
Issue number5
DOIs
Publication statusPublished - Jan 1 2001
Externally publishedYes

Fingerprint

Reparametrization
learning
Probability Distribution
formulations
pattern recognition
invariance
Formulation
Pattern Recognition
Learning Algorithm
Invariance
Gauge
Learning

ASJC Scopus subject areas

  • Statistical and Nonlinear Physics
  • Statistics and Probability
  • Condensed Matter Physics

Cite this

@article{9483745214304d47b2af82f66059b5d9,
title = "Reparametrization-covariant theory for on-line learning of probability distributions",
abstract = "We discuss the on-line learning of probability distributions in a reparametrization covariant formulation. Reparametrization covariance plays an essential role not only to respect an intrinsic property of “information” but also for pattern recognition problems. We can obtain an optimal on-line learning algorithm with reparametrization invariance, where the conformal gauge connects a covariant formulation with a noncovariant one in a natural way.",
author = "Toshiaki Aida",
year = "2001",
month = "1",
day = "1",
doi = "10.1103/PhysRevE.64.056128",
language = "English",
volume = "64",
journal = "Physical Review E",
issn = "2470-0045",
publisher = "American Physical Society",
number = "5",

}

TY - JOUR

T1 - Reparametrization-covariant theory for on-line learning of probability distributions

AU - Aida, Toshiaki

PY - 2001/1/1

Y1 - 2001/1/1

N2 - We discuss the on-line learning of probability distributions in a reparametrization covariant formulation. Reparametrization covariance plays an essential role not only to respect an intrinsic property of “information” but also for pattern recognition problems. We can obtain an optimal on-line learning algorithm with reparametrization invariance, where the conformal gauge connects a covariant formulation with a noncovariant one in a natural way.

AB - We discuss the on-line learning of probability distributions in a reparametrization covariant formulation. Reparametrization covariance plays an essential role not only to respect an intrinsic property of “information” but also for pattern recognition problems. We can obtain an optimal on-line learning algorithm with reparametrization invariance, where the conformal gauge connects a covariant formulation with a noncovariant one in a natural way.

UR - http://www.scopus.com/inward/record.url?scp=85035264800&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85035264800&partnerID=8YFLogxK

U2 - 10.1103/PhysRevE.64.056128

DO - 10.1103/PhysRevE.64.056128

M3 - Article

AN - SCOPUS:85035264800

VL - 64

JO - Physical Review E

JF - Physical Review E

SN - 2470-0045

IS - 5

ER -