Box-simplex games are a family of bilinear minimax objectives which encapsulate graph-structured problems such as maximum flow [She17], optimal transport [JST19], and bipartite matching [AJJ+22]. We develop efficient near-linear time, high-accuracy solvers for regularized variants of these games. Beyond the immediate applications of such solvers for computing Sinkhorn distances, a prominent tool in machine learning, we show that these solvers can be used to obtain improved running times for maintaining a (fractional) $\epsilon$-approximate maximum matching in a dynamic decremental bipartite graph against an adaptive adversary. We give a generic framework which reduces this dynamic matching problem to solving regularized graph-structured optimization problems to high accuracy. Through our reduction framework, our regularized box-simplex game solver implies a new algorithm for dynamic decremental bipartite matching in total time $\tilde{O}(m \cdot \epsilon^{-3})$, from an initial graph with $m$ edges and $n$ nodes. We further show how to use recent advances in flow optimization [CKL+22] to improve our runtime to $m^{1 + o(1)} \cdot \epsilon^{-2}$, thereby demonstrating the versatility of our reduction-based approach. These results improve upon the previous best runtime of $\tilde{O}(m \cdot \epsilon^{-4})$ [BGS20] and illustrate the utility of using regularized optimization problem solvers for designing dynamic algorithms.
翻译:Box-Semplex 游戏是双线小型移动目标的组合, 它包含最大流 [She17] 、 最佳双部分迁移 [JST19] 和双部分匹配 [AJJ+22] 等图形结构问题。 我们开发了高效的近线时间, 对这些游戏的常规变体开发了高精度解答器。 除了这些解答器直接用于计算 Sinkhorn 距离之外, 一个突出的机器学习工具, 我们显示这些解答器可以用来获得更好的运行时间, 用于维持一个( 折数) $\epsilon$- 的( 折数) 的( 折数) 基数( 折数) 、 优化双部分运量( JST19 ) 和双部分对齐匹配。 我们给出了一个通用框架, 将这种动态匹配问题降低到高精度。 通过我们的减框架, 我们固定的框- 简单解算器解算器解算器意味着新的双部分匹配时间 ${O} (m\\\ dload=xxxxx romodeal rus prus prent prus pal_ listral=xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx