Kernekoncepter
Improving worst-case approximation ratios for the MaxCut problem using noisy and partial predictions.
Resumé
グラフカット問題の近似可能性を、ノイズのある予測と部分的な予測を使用して改善する方法に焦点を当てた研究。ノイジーな予測モデルと部分的な予測モデルに基づいて、最悪ケースの近似比率を向上させるアルゴリズムが提案されています。これらの手法は、最大カット問題における近似比率を改善し、理論と実践の間のギャップを埋めることができます。
Statistik
α ≈ 0.878 is the MaxCut threshold.
β ≈ 0.858 is the approximation ratio for MaxBisection.
ε4, ε, Ω(ε), Ω(ε2) are used to quantify improvements in approximation ratios.
Citater
"In recent years, the abundance of data and the impact of machine learning has led to algorithmic models that seek to go beyond worst-case performance using a noisy prediction of an optimal solution."
"Motivated by this vision, in his SODA ’23 plenary lecture, Ola Svensson posed the following question: In the MaxCut problem, suppose we are given a prediction for the optimal cut that is independently correct for every vertex with probability 1/2 + ε."
"We show how these predictions can be used to improve on the worst-case approximation ratios for this problem."