Machine Learning and Risk Management: SVDD Meets RQE

Georges Dionne () and Gilles Boevi Koumou ()
Additional contact information
Gilles Boevi Koumou: HEC Montreal, Canada Research Chair in Risk Management

No 18-6, Working Papers from HEC Montreal, Canada Research Chair in Risk Management

Abstract: This paper re-examines Rao’s Quadratic Entropy (RQE) portfolio diversification in the light of the support vector data description (SVDD), an unsupervised machine learning algorithm developed for the one-class classification. It makes the link between RQE portfolio diversification and the SVDD. More specifically, we show that RQE portfolio diversification problem is equivalent to the dual representation of a hard-margin SVDD problem. This result demonstrates, on the one hand, that the SVDD and its rich set of extensions are relevant for risk management, in particular portfolio diversification. On the second hand, it provides new insights on RQE portfolio diversification approach in terms of interpretation, understanding, implementation, specification, and optimization. This strengthens the believe that machine learning can play an important role in risk management and RQE is an adequate framework to quantify and to manage portfolio diversification.

Keywords: Machine Learning; One-Class Classification; Support Vector Data Description; Rao’s Quadratic Entropy; Portfolio Diversification (search for similar items in EconPapers)
JEL-codes: D83 G11 (search for similar items in EconPapers)
Pages: 36 pages
Date: 2018-11-30
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.risksresearch.com/_files/ugd/a6eed3_85 ... 99a071a37798cfc8.pdf Full text (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:ris:crcrmw:2018_006

Access Statistics for this paper

More papers in Working Papers from HEC Montreal, Canada Research Chair in Risk Management Contact information at EDIRC.
Bibliographic data for series maintained by Claire Boisvert ().