Regime Aware Learning

Marcus Bendtsen
; Proceedings of the Eighth International Conference on Probabilistic Graphical Models, PMLR 52:1-12, 2016.

Abstract

We propose a regime aware learning algorithm to learn a sequence of Bayesian networks (BNs) that model a system that undergoes \it 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.

Cite this Paper


BibTeX
@InProceedings{pmlr-v52-bendtsen16, title = {Regime Aware Learning}, author = {Marcus Bendtsen}, booktitle = {Proceedings of the Eighth International Conference on Probabilistic Graphical Models}, pages = {1--12}, year = {2016}, editor = {Alessandro Antonucci and Giorgio Corani and Cassio Polpo Campos}}, volume = {52}, series = {Proceedings of Machine Learning Research}, address = {Lugano, Switzerland}, month = {06--09 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v52/bendtsen16.pdf}, url = {http://proceedings.mlr.press/v52/bendtsen16.html}, abstract = {We propose a regime aware learning algorithm to learn a sequence of Bayesian networks (BNs) that model a system that undergoes \it 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.} }
Endnote
%0 Conference Paper %T Regime Aware Learning %A Marcus Bendtsen %B Proceedings of the Eighth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2016 %E Alessandro Antonucci %E Giorgio Corani %E Cassio Polpo Campos} %F pmlr-v52-bendtsen16 %I PMLR %J Proceedings of Machine Learning Research %P 1--12 %U http://proceedings.mlr.press %V 52 %W PMLR %X We propose a regime aware learning algorithm to learn a sequence of Bayesian networks (BNs) that model a system that undergoes \it 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.
RIS
TY - CPAPER TI - Regime Aware Learning AU - Marcus Bendtsen BT - Proceedings of the Eighth International Conference on Probabilistic Graphical Models PY - 2016/08/15 DA - 2016/08/15 ED - Alessandro Antonucci ED - Giorgio Corani ED - Cassio Polpo Campos} ID - pmlr-v52-bendtsen16 PB - PMLR SP - 1 DP - PMLR EP - 12 L1 - http://proceedings.mlr.press/v52/bendtsen16.pdf UR - http://proceedings.mlr.press/v52/bendtsen16.html AB - We propose a regime aware learning algorithm to learn a sequence of Bayesian networks (BNs) that model a system that undergoes \it 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. ER -
APA
Bendtsen, M.. (2016). Regime Aware Learning. Proceedings of the Eighth International Conference on Probabilistic Graphical Models, in PMLR 52:1-12

Related Material