My main research interests are approximation algorithms for combinatorial optimization problems, in particular, primal-dual approaches and network design.
Publications
The Bidirected Cut Relaxation for Steiner Tree: Better Integrality Gap Bounds and the Limits of Moat Growing with V. Traub [arXiv]
The main technique used to bound the integrality gap of BCR is a generalization of moat growing to a (bi)directed setting which was introduced here by Byrka, Grandoni, and Traub.
Bidirecred moat growing can be quite chaotic and unintuitive, which is why we provide animated versions of the figures in our paper: