Core Concepts
The authors explore Max-Cut approximations using noisy and partial predictions, achieving significant improvements over worst-case scenarios.
Abstract
The content delves into the study of Max-Cut problems under various prediction models. It introduces algorithms for both noisy and partial predictions, showcasing advancements in approximation ratios. The research extends to general 2-CSPs, demonstrating the efficacy of these methods across different problem instances.
Stats
We give an algorithm that achieves an α + eΩ(ε4)-approximation for the noisy predictions model.
We can also give a β + Ω(ε)-approximation for the partial predictions model.
For low-degree graphs, we use the algorithm by Feige, Karpinski, and Langberg to guarantee an (αGW+ eO(1/d2))-approximation where d is the maximum degree.
Quotes
"We show how these predictions can be used to improve on the worst-case approximation ratios for this problem."
"This paradigm has been particularly successful at overcoming information-theoretic barriers in online algorithms."