Communications of the ACM
Research Archive
The Research archive provides access to all Research articles published in past issues of Communications of the ACM.
February 2023
From Communications of the ACM
Proving Data-Poisoning Robustness in Decision Trees
We present a sound verification technique based on abstract interpretation and implement it in a tool called Antidote, which abstractly trains decision trees for an intractably large space of possible poisoned datasets.
From Communications of the ACM
Technical Perspective: Beautiful Symbolic Abstractions for Safe and Secure Machine Learning
"Proving Data-Poisoning Robustness in Decision Trees," by Samuel Drews et al., addresses the challenge of processing an intractably large set of trained models when enumeration is infeasible in a clean, beautiful, and elegant…