English

Mirrorless Mirror Descent: A Natural Derivation of Mirror Descent

Machine Learning 2021-07-05 v3 Optimization and Control Machine Learning

Abstract

We present a primal only derivation of Mirror Descent as a "partial" discretization of gradient flow on a Riemannian manifold where the metric tensor is the Hessian of the Mirror Descent potential. We contrast this discretization to Natural Gradient Descent, which is obtained by a "full" forward Euler discretization. This view helps shed light on the relationship between the methods and allows generalizing Mirror Descent to general Riemannian geometries, even when the metric tensor is {\em not} a Hessian, and thus there is no "dual."

Keywords

Cite

@article{arxiv.2004.01025,
  title  = {Mirrorless Mirror Descent: A Natural Derivation of Mirror Descent},
  author = {Suriya Gunasekar and Blake Woodworth and Nathan Srebro},
  journal= {arXiv preprint arXiv:2004.01025},
  year   = {2021}
}

Comments

11 pages