BriefGPT.xyz
May, 2023
从数据中学习具有少量根因的有向无环图（DAG）
Learning DAGs from Data with Few Root Causes
HTML
PDF
Panagiotis Misiakos, Chris Wendler, Markus Püschel
TL;DR
提出了一种新的算法, 从线性结构方程模型（SEM）生成的数据中学习有向无环图（DAG）。该算法证明在新设置中的可识别性，并显示与之前的DAG学习方法相比，少量有噪声的root causes能提供优异性能。
Abstract
We present a novel perspective and algorithm for learning
directed acyclic graphs
(DAGs) from data generated by a
linear structural equation model
(SEM). First, we show that a linear SEM can be viewed as a linear
→