发帖
雷达卡
Efficient Training of Robust Decision Trees Against Adversarial Examples
Daniel Vos 1 Sicco Verwer 1
Abstract et al., 2019), we closely mimic the greedy recursive split-
ting strategy that traditional decision trees use and we score
Current state-of-the-art algorithms for training ro-
splits with the adversarial Gini impurity. We prove that
bust decision trees have high run ...
京公网安备 11010802022788号
