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 2004.00668cs.LGstat.ML
收藏

利用附加重要性度量了解全局特征贡献

Understanding Global Feature Contributions With Additive Importance Measures

University of Washington(华盛顿大学);Microsoft Research(微软研究院)

Ian Covert, Scott Lundberg, Su-In Lee

摘要:

理解复杂机器学习模型的内部工作原理是一个长期存在的问题,最近的研究主要集中在局部可解释性上。为了从全局意义上评估单个输入特征的作用,我们探索了通过与每个特征相关联的预测能力来定义特征重要性的观点。我们引入了两个概念的预测能力(基于模型的和通用的),并正式这种方法的框架的添加剂的重要性措施,它统一了众多的方法在文献中。然后,我们提出了SAGE,一个模型无关的方法,量化预测能力,同时考虑功能的相互作用。我们的实验表明,SAGE可以有效地计算,它分配更准确的重要性值比其他方法。

英文摘要:

Understanding the inner workings of complex machine learning models is a long-standing problem and most recent research has focused on local interpretability. To assess the role of individual input features in a global sense, we explore the perspective of defining feature importance through the predictive power associated with each feature. We introduce two notions of predictive power (model-based and universal) and formalize this approach with a framework of additive importance measures, which unifies numerous methods in the literature. We then propose SAGE, a model-agnostic method that quantifies predictive power while accounting for feature interactions. Our experiments show that SAGE can be calculated efficiently and that it assigns more accurate importance values than other methods.

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。

↑