Adapting Bayes Network Structures to Non-stationary Domains

Søren Holbech Nielsen, Thomas Dyhre Nielsen

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

6 Citations (Scopus)

Abstract

When an incremental structural learning method gradually modifies a Bayesian network (BN) structure to fit observations, as they are read from a database, we call the process structural adaptation. Structural adaptation is useful when the learner is set to work in an unknown environment, where a BN is to be gradually constructed as observations of the environment are made. Existing algorithms for incremental learning assume that the samples in the database have been drawn from a single underlying distribution. In this paper we relax this assumption, so that the underlying distribution can change during the sampling of the database. The method that we present can thus be used in unknown environments, where it is not even known whether the dynamics of the environment are stable. We briefly state formal correctness results for our method, and demonstrate its feasibility experimentally.
Original languageEnglish
Title of host publicationProceedings of the Third European Workshop on Probabilistic Graphical Models
Number of pages8
Publication date2006
Pages223-230
Publication statusPublished - 2006
EventEuropean Workshop on Probabilistic Graphical Models - Prag, Czech Republic
Duration: 12 Sep 200615 Sep 2006
Conference number: 3

Conference

ConferenceEuropean Workshop on Probabilistic Graphical Models
Number3
CountryCzech Republic
CityPrag
Period12/09/200615/09/2006

Fingerprint

Bayesian networks
Sampling

Keywords

  • Bayesian networks
  • Adaptation
  • Belief revision

Cite this

Nielsen, S. H., & Nielsen, T. D. (2006). Adapting Bayes Network Structures to Non-stationary Domains. In Proceedings of the Third European Workshop on Probabilistic Graphical Models (pp. 223-230)
Nielsen, Søren Holbech ; Nielsen, Thomas Dyhre. / Adapting Bayes Network Structures to Non-stationary Domains. Proceedings of the Third European Workshop on Probabilistic Graphical Models. 2006. pp. 223-230
@inproceedings{3cbe0650a0db11db8ed6000ea68e967b,
title = "Adapting Bayes Network Structures to Non-stationary Domains",
abstract = "When an incremental structural learning method gradually modifies a Bayesian network (BN) structure to fit observations, as they are read from a database, we call the process structural adaptation. Structural adaptation is useful when the learner is set to work in an unknown environment, where a BN is to be gradually constructed as observations of the environment are made. Existing algorithms for incremental learning assume that the samples in the database have been drawn from a single underlying distribution. In this paper we relax this assumption, so that the underlying distribution can change during the sampling of the database. The method that we present can thus be used in unknown environments, where it is not even known whether the dynamics of the environment are stable. We briefly state formal correctness results for our method, and demonstrate its feasibility experimentally.",
keywords = "Bayesian networks, Adaptation, Belief revision",
author = "Nielsen, {S{\o}ren Holbech} and Nielsen, {Thomas Dyhre}",
year = "2006",
language = "English",
pages = "223--230",
booktitle = "Proceedings of the Third European Workshop on Probabilistic Graphical Models",

}

Nielsen, SH & Nielsen, TD 2006, Adapting Bayes Network Structures to Non-stationary Domains. in Proceedings of the Third European Workshop on Probabilistic Graphical Models. pp. 223-230, Prag, Czech Republic, 12/09/2006.

Adapting Bayes Network Structures to Non-stationary Domains. / Nielsen, Søren Holbech; Nielsen, Thomas Dyhre.

Proceedings of the Third European Workshop on Probabilistic Graphical Models. 2006. p. 223-230.

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Adapting Bayes Network Structures to Non-stationary Domains

AU - Nielsen, Søren Holbech

AU - Nielsen, Thomas Dyhre

PY - 2006

Y1 - 2006

N2 - When an incremental structural learning method gradually modifies a Bayesian network (BN) structure to fit observations, as they are read from a database, we call the process structural adaptation. Structural adaptation is useful when the learner is set to work in an unknown environment, where a BN is to be gradually constructed as observations of the environment are made. Existing algorithms for incremental learning assume that the samples in the database have been drawn from a single underlying distribution. In this paper we relax this assumption, so that the underlying distribution can change during the sampling of the database. The method that we present can thus be used in unknown environments, where it is not even known whether the dynamics of the environment are stable. We briefly state formal correctness results for our method, and demonstrate its feasibility experimentally.

AB - When an incremental structural learning method gradually modifies a Bayesian network (BN) structure to fit observations, as they are read from a database, we call the process structural adaptation. Structural adaptation is useful when the learner is set to work in an unknown environment, where a BN is to be gradually constructed as observations of the environment are made. Existing algorithms for incremental learning assume that the samples in the database have been drawn from a single underlying distribution. In this paper we relax this assumption, so that the underlying distribution can change during the sampling of the database. The method that we present can thus be used in unknown environments, where it is not even known whether the dynamics of the environment are stable. We briefly state formal correctness results for our method, and demonstrate its feasibility experimentally.

KW - Bayesian networks

KW - Adaptation

KW - Belief revision

M3 - Article in proceeding

SP - 223

EP - 230

BT - Proceedings of the Third European Workshop on Probabilistic Graphical Models

ER -

Nielsen SH, Nielsen TD. Adapting Bayes Network Structures to Non-stationary Domains. In Proceedings of the Third European Workshop on Probabilistic Graphical Models. 2006. p. 223-230