BriefGPT.xyz
Sep, 2023
无约束结构学习与平滑无环定向
Constraint-Free Structure Learning with Smooth Acyclic Orientations
HTML
PDF
Riccardo Massidda, Francesco Landolfi, Martina Cinquini, Davide Bacciu
TL;DR
COSMO是一种无约束的连续优化方案,用于无环结构学习,通过参数化的优先级向量定义可微分的方向矩阵的近似,能够以渐近更快的速度收敛到无环解,且在图结构重建方面相比其他方法表现优异。
Abstract
The
structure learning
problem consists of fitting data generated by a
directed acyclic graph
(DAG) to correctly reconstruct its arcs. In this context,
→