cs.LGNov 12, 2026

A Convex-Analytic Reformulation for Poisoning Attacks on Logistic Regression

Authors: Carina Moreira Costa, Maria Eduarda Pinheiro, Martin Schmidt

Abstract

We study data poisoning attacks on logistic regression with ℓ2\ell_2-regularization, where an attacker chooses the features of a fixed number of injected observations with prescribed labels to maximize the logistic loss on the clean training data. Strong convexity of the defender's problem yields an exact reformulation through its first-order optimality conditions. Using convex conjugacy of the softplus function, we replace the sigmoid relations by a single convex optimality-gap inequality and express the attacker's objective as a concave function. A change of variables makes the stationarity condition affine, leaving bounded bilinear equalities as the only nonconvex constraints. We derive finite bounds on all variables involved in these equalities, including a global bound on the coefficient norm and tighter bounds obtained by preprocessing based on convex optimization. Experiments on eight binary classification datasets evaluate attack effectiveness and compare the two reformulations when solved with SCIP. Attacks obtained by local optimization degrade predictive performance on most datasets, while the convex-analytic formulation produces substantially stronger reported dual bounds within a fixed time limit.

Explore similar work

CardsList
  1. Exact Certification of Data-Poisoning Attacks Using Mixed-Integer Programming

    Feb 18, 2026Philip Sosnin, Jodie Knapp, Fraser Kennedy +2Data Poisoning AttacksCertified Robustness

  2. Convex training of Lipschitz-regularized shallow neural networks

    Jun 17, 2026Chao Yin, Antoine Lesage-LandryAdversarial TrainingShallow Neural Networks