The current best practice for computing optimal transport (OT) is via entropy regularization and Sinkhorn iterations. This algorithm runs in quadratic time as it requires the full pairwise cost matrix, which is prohibitively expensive for large sets of objects. In this work we propose two effective log-linear time approximations of the cost matrix: First, a sparse approximation based on locality-sensitive hashing (LSH) and, second, a Nystr\"om approximation with LSH-based sparse corrections, which we call locally corrected Nystr\"om (LCN). These approximations enable general log-linear time algorithms for entropy-regularized OT that perform well even for the complex, high-dimensional spaces common in deep learning. We analyse these approximations theoretically and evaluate them experimentally both directly and end-to-end as a component for real-world applications. Using our approximations for unsupervised word embedding alignment enables us to speed up a state-of-the-art method by a factor of 3 while also improving the accuracy by 3.1 percentage points without any additional model changes. For graph distance regression we propose the graph transport network (GTN), which combines graph neural networks (GNNs) with enhanced Sinkhorn. GTN outcompetes previous models by 48% and still scales log-linearly in the number of nodes.
翻译:计算最佳运输( OT) 的最佳方法( OT) 是通过 entropy 正规化和 Sinkhorn 迭代。 这种算法在四进制时间运行, 因为它需要全对对称成本矩阵, 这对于大型天体来说是极其昂贵的。 在这项工作中, 我们提议了两种有效的成本矩阵日志线时间近似值: 首先, 基于地点敏感散列( LSH) 和基于 LSH 的基于 LSH 的稀散校正( 我们称之为本地校正 Nystr\"om (LCN) ) 的微缩缩缩缩缩略图。 这些近似法可以在四进式OTA 上进行一般的对正对线时间算算算算算, 即使对于复杂、 高维的空格空间也是如此。 我们从理论上分析这些近似时间近似值, 并直接和端到端端到端对成本应用进行实验。 使用我们的近似的单词嵌嵌嵌化词的缩缩缩缩缩缩略图, 使我们以3 速度加快一个状态方法, 同时将精确度提高3. 1 1% 的 的 模式 。