论文标题
反伴侣神经建筑搜索模型防御
Anti-Bandit Neural Architecture Search for Model Defense
论文作者
论文摘要
深度卷积神经网络（DCNN）已成为机器学习中表现最好的人，但可以受到对抗性攻击的挑战。在本文中，我们使用神经体系结构搜索（NAS）来防止对抗性攻击，该搜索是基于对denoising块，无举重操作，Gabor过滤器和卷积的全面搜索。由此产生的反伴侣NAS（Abanditnas）基于上下置信度范围（LCB和UCB）结合了一种新的操作评估措施和搜索过程。与仅使用UCB进行评估的传统匪徒算法不同，我们使用UCB放弃武器以搜索效率，而LCB则在手臂之间进行公平的竞争。广泛的实验表明，Abanditnas比其他NAS方法快，而在PGD- $ 7 $下获得了$ 8.73 \％$的改善。
Deep convolutional neural networks (DCNNs) have dominated as the best performers in machine learning, but can be challenged by adversarial attacks. In this paper, we defend against adversarial attacks using neural architecture search (NAS) which is based on a comprehensive search of denoising blocks, weight-free operations, Gabor filters and convolutions. The resulting anti-bandit NAS (ABanditNAS) incorporates a new operation evaluation measure and search process based on the lower and upper confidence bounds (LCB and UCB). Unlike the conventional bandit algorithm using UCB for evaluation only, we use UCB to abandon arms for search efficiency and LCB for a fair competition between arms. Extensive experiments demonstrate that ABanditNAS is faster than other NAS methods, while achieving an $8.73\%$ improvement over prior arts on CIFAR-10 under PGD-$7$.
