28 Feb 2024
Algorithms are presented that provably exceed the Goemans-Williamson approximation ratio for the NP-hard Max-Cut problem by incorporating imperfect prediction information, with specific improvements of Ω(ε⁴) for noisy predictions and Ω(ε) for partial predictions. The techniques also extend to dense 2-CSPs.
There are no more papers matching your filters at the moment.
