English

cCorrGAN: Conditional Correlation GAN for Learning Empirical Conditional Distributions in the Elliptope

Statistical Finance 2021-07-23 v1 Machine Learning

Abstract

We propose a methodology to approximate conditional distributions in the elliptope of correlation matrices based on conditional generative adversarial networks. We illustrate the methodology with an application from quantitative finance: Monte Carlo simulations of correlated returns to compare risk-based portfolio construction methods. Finally, we discuss about current limitations and advocate for further exploration of the elliptope geometry to improve results.

Keywords

Cite

@article{arxiv.2107.10606,
  title  = {cCorrGAN: Conditional Correlation GAN for Learning Empirical Conditional Distributions in the Elliptope},
  author = {Gautier Marti and Victor Goubet and Frank Nielsen},
  journal= {arXiv preprint arXiv:2107.10606},
  year   = {2021}
}

Comments

International Conference on Geometric Science of Information

R2 v1 2026-06-24T04:25:38.415Z