.png){kind=icon}
返回
Reducing classifier overconfidence against adversaries through graph algorithms
.png){kind=icon}
OA
AI
DOI：10.1007/s10994-023-06307-y.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
In this work we show that deep learning classifiers tend to become overconfident in their answers under adversarial attacks, even when the classifier is optimized to survive such attacks. Our work draws upon stochastic geometry and graph algorithms to propose a general framework to replace the last fully connected layer and softmax output. This framework (a) can be applied to any classifier and (b) significantly reduces the classifier's overconfidence in its output without much of an impact on its accuracy when compared to original adversarially-trained classifiers. Its relative effectiveness increases as the attacker becomes more powerful. Our use of graph algorithms in adversarial learning is new and of independent interest. Finally, we show the advantages of this last-layer softmax replacement over image tasks under common adversarial attacks.
Keyword:
Adversarial robustness
Overconfidence
Gossip algorithm
期刊
IF:
2.9
论文数:
2.7K
被引数:
3.4W
机构
引用论文
.png){kind=icon}
Welding characteristics of aluminum, copper, nickel and aluminum alloy with alumina coating using ultrasonic complex vibration welding equipments铝、铜、镍及铝合金氧化铝涂层超声复合振动焊接特性研究
PREAI
分享收藏分享收藏分享收藏
查看更多内容