liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Regime Aware Learning
Linköping University, Department of Computer and Information Science, Database and information techniques. Linköping University, Faculty of Science & Engineering.ORCID iD: 0000-0002-8678-1164
2016 (English)In: Proceedings of the Eighth International Conference on Probabilistic Graphical Models, JMLR , 2016, Vol. 52, 1-12 p.Conference paper (Refereed)
Abstract [en]

We propose a regime aware learning algorithm to learn a sequence of Bayesian networks (BNs) that model a system that undergoes regime changes. The last BN in the sequence represents the system’s current regime, and should be used for BN inference. To explore the feasibility of the algorithm, we create baseline tests against learning a singe BN, and show that our proposed algorithm outperforms the single BN approach. We also apply the learning algorithm on real world data from the financial domain, where it is evident that the algorithm is able to produce BNs that have adapted to the regime changes during the most recent global financial crisis of 2007-08.

Place, publisher, year, edition, pages
JMLR , 2016. Vol. 52, 1-12 p.
Series
, JMLR: Workshop and Conference Proceedings, ISSN 1938-7228 ; 52
Keyword [en]
Regime changes, Bayesian networks, financial application
National Category
Computer and Information Science
Identifiers
URN: urn:nbn:se:liu:diva-131133OAI: oai:DiVA.org:liu-131133DiVA: diva2:968120
Conference
Eighth International Conference on Probabilistic Graphical Models (PGM 2016), Lugano, Switzerland, September 6-9, 2016
Available from: 2016-09-12 Created: 2016-09-12 Last updated: 2016-09-19Bibliographically approved

Open Access in DiVA

No full text

Other links

Link to publication

Search in DiVA

By author/editor
Bendtsen, Marcus
By organisation
Database and information techniquesFaculty of Science & Engineering
Computer and Information Science

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 30 hits
ReferencesLink to record
Permanent link

Direct link