学习DAG而不强加非周期性
Learning DAGs without imposing acyclicity
摘要：
我们探讨，如果它是可能的学习有向无环图（DAG）的数据，而不明确施加的无环约束。特别是，对于高斯分布，我们框架的结构学习作为一个稀疏矩阵分解问题，我们的经验表明，解决一个$\ell_1$ -惩罚优化收益率良好的恢复的真实图，并在一般情况下，几乎DAG图。此外，这种方法是计算效率高，是不受爆炸的组合复杂性，在经典的结构学习算法。
英文摘要：
We explore if it is possible to learn a directed acyclic graph (DAG) from data without imposing explicitly the acyclicity constraint. In particular, for Gaussian distributions, we frame structural learning as a sparse matrix factorization problem and we empirically show that solving an $\ell_1$-penalized optimization yields to good recovery of the true graph and, in general, to almost-DAG graphs. Moreover, this approach is computationally efficient and is not affected by the explosion of combinatorial complexity as in classical structural learning algorithms.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
