arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.05660cs.LGcs.PLstat.ML
收藏

抽象神经网络

Abstract Neural Networks

University of California, Davis(加州大学戴维斯分校)

Matthew Sotoudeh, Aditya V. Thakur

摘要:

深度神经网络(DNN)正迅速应用于无人机和飞机控制等安全关键领域,激发了验证其行为安全性的技术。不幸的是,DNN验证是NP难的,当前的算法随着DNN中节点的数量呈指数级增长。本文介绍了抽象神经网络(ANN)的概念,它可以用来在使用较少节点的情况下完全过度逼近DNN。ANN类似于DNN,除了权重矩阵被给定抽象域中的值所取代。我们提出了一个由抽象域和DNN中使用的激活函数参数化的框架,可以用来构建相应的ANN。我们提出了DNN激活函数的必要和充分条件,以使所构建的ANN完全过逼近给定的DNN。先前关于DNN抽象的工作仅限于区间域和ReLU激活函数。我们的框架可以用其他抽象域(如八边形和多面体)以及其他激活函数(如Leaky ReLU、Sigmoid和Hyperbolic Tangent)来实例化。

英文摘要:

Deep Neural Networks (DNNs) are rapidly being applied to safety-critical domains such as drone and airplane control, motivating techniques for verifying the safety of their behavior. Unfortunately, DNN verification is NP-hard, with current algorithms slowing exponentially with the number of nodes in the DNN. This paper introduces the notion of Abstract Neural Networks (ANNs), which can be used to soundly overapproximate DNNs while using fewer nodes. An ANN is like a DNN except weight matrices are replaced by values in a given abstract domain. We present a framework parameterized by the abstract domain and activation functions used in the DNN that can be used to construct a corresponding ANN. We present necessary and sufficient conditions on the DNN activation functions for the constructed ANN to soundly over-approximate the given DNN. Prior work on DNN abstraction was restricted to the interval domain and ReLU activation function. Our framework can be instantiated with other abstract domains such as octagons and polyhedra, as well as other activation functions such as Leaky ReLU, Sigmoid, and Hyperbolic Tangent.

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。

↑