arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2010.06175stat.MLcs.LG
收藏

用于神经网络控制特征选择的神经高斯镜

Neural Gaussian Mirror for Controlled Feature Selection in Neural Networks

Department of Statistics, Virginia Tech(弗吉尼亚理工大学统计系);Department of Statistics, University of Chicago(芝加哥大学统计系);Department of Statistics, Harvard University(哈佛大学统计系)

Xin Xing, Yu Gui, Chenguang Dai, Jun S. Liu

摘要:

深度神经网络(DNN)越来越受欢迎,并在预测任务中取得了出色的表现。然而,DNN框架本身无法告知用户哪些特征与预测或多或少相关,这限制了其在许多科学领域的适用性。我们引入神经高斯镜(NGM),其中镜像功能的创建,通过基于核的条件依赖度量的结构化扰动,以帮助评估功能的重要性。我们设计了DNN架构的两个修改,用于合并镜像功能并提供镜像统计来衡量功能的重要性。如在模拟和真实数据的例子中所示,所提出的方法控制在预定义的水平的特征选择错误率,并保持高的选择功率,即使存在高度相关的功能。

英文摘要:

Deep neural networks (DNNs) have become increasingly popular and achieved outstanding performance in predictive tasks. However, the DNN framework itself cannot inform the user which features are more or less relevant for making the prediction, which limits its applicability in many scientific fields. We introduce neural Gaussian mirrors (NGMs), in which mirrored features are created, via a structured perturbation based on a kernel-based conditional dependence measure, to help evaluate feature importance. We design two modifications of the DNN architecture for incorporating mirrored features and providing mirror statistics to measure feature importance. As shown in simulated and real data examples, the proposed method controls the feature selection error rate at a predefined level and maintains a high selection power even with the presence of highly correlated features.

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。

↑