Unifying strongly polynomial algorithms for subclasses of Linear Programs

Published on ● Video Link: https://www.youtube.com/watch?v=53Iw0kufukQ



Duration: 11:06
149 views
4


Bento Natura (Georgia Tech)
https://simons.berkeley.edu/talks/bento-natura-georgia-tech-2023-09-08
Meet the Fellows Welcome Event Fall 2023

The existence of a strongly polynomial algorithm for linear programs (LP) is one of the most important open problems in theoretical computer science. The last decades have seen tremendous progress for special subclasses of LP (e.g., minimum cost flow, multi-commodity flow, discounted Markov decision processes) whose proofs of strong polynomiality rely on very different ideas, algorithms and techniques. Together with Allamigeon, Dadush, Loho, and Végh (FOCS 2022) we devised an interior point method that runs in strongly polynomial time whenever a certain curvature is polynomially bounded. Taking this algorithm as a blackbox, we show how strong polynomiality for many of the subclasses of LP can be recovered.







Tags:
Simons Institute
theoretical computer science
UC Berkeley
Computer Science
Theory of Computation
Theory of Computing
Meet the Fellows Welcome Event Fall 2023
Bento Natura