arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2106.02835cs.LGcs.AI
收藏

基于熵损失在连续优化中学习因果结构的作用研究

On the Role of Entropy-based Loss for Learning Causal Structures with Continuous Optimization

Guangdong University of Technology(广东工业大学);Peng Cheng Laboratory(鹏城实验室);Shantou University(汕头大学)

Weilin Chen, Jie Qiao, Ruichu Cai, Zhifeng Hao

AI总结:

本文针对NOTEARS方法在高斯噪声假设不成立时的局限,提出一种基于熵的通用损失函数,在连续优化框架下提升因果结构学习性能,并在多项指标上取得最优结果。

摘要:

从观测数据中进行因果发现是许多科学领域中一项重要但具有挑战性的任务。最近,一种名为NOTEARS的具有非组合有向无环约束的方法,将因果结构学习问题表述为使用最小二乘损失的连续优化问题。尽管最小二乘损失函数在标准高斯噪声假设下具有良好的合理性,但当该假设不成立时,其适用性受到限制。在这项工作中,我们从理论上证明,违反高斯噪声假设将阻碍因果方向的识别,使得在线性情况下因果方向完全由因果强度以及噪声方差决定,而在非线性情况下则由强非高斯噪声决定。因此,我们提出了一种更通用的基于熵的损失函数,该损失函数在理论上与任意噪声分布下的似然得分一致。我们在合成数据和真实世界数据上进行了广泛的实证评估,以验证所提方法的有效性,并表明我们的方法在结构汉明距离、错误发现率和真阳性率指标上取得了最佳结果。

英文摘要:

Causal discovery from observational data is an important but challenging task in many scientific fields. Recently, a method with non-combinatorial directed acyclic constraint, called NOTEARS, formulates the causal structure learning problem as a continuous optimization problem using least-square loss. Though the least-square loss function is well justified under the standard Gaussian noise assumption, it is limited if the assumption does not hold. In this work, we theoretically show that the violation of the Gaussian noise assumption will hinder the causal direction identification, making the causal orientation fully determined by the causal strength as well as the variances of noises in the linear case and by the strong non-Gaussian noises in the nonlinear case. Consequently, we propose a more general entropy-based loss that is theoretically consistent with the likelihood score under any noise distribution. We run extensive empirical evaluations on both synthetic data and real-world data to validate the effectiveness of the proposed method and show that our method achieves the best in Structure Hamming Distance, False Discovery Rate, and True Positive Rate matrices.

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。

↑