BriefGPT.xyz
Jun, 2024
神经网络中S形非线性函数的线性界限函数自动设计
Automated Design of Linear Bounding Functions for Sigmoidal Nonlinearities in Neural Networks
HTML
PDF
Matthias König, Xiyue Zhang, Holger H. Hoos, Marta Kwiatkowska, Jan N. van Rijn
TL;DR
通过使用一种简单的搜索方法,精心地根据最先进的算法配置技术调整给定的验证问题,我们提出了一种新颖的参数搜索方法来改进这些线性逼近的质量,进而在几个常用的本地鲁棒性验证基准上平均提高了25%的全局下界。
Abstract
The ubiquity of
deep learning algorithms
in various applications has amplified the need for assuring their
robustness
against small input perturbations such as those occurring in
→