Structural Causal Models Reveal Confounder Bias in Linear Program Modelling

TL;DR

This paper explores how structural causal models can reveal confounder bias in linear programming, showing that adversarial-like attacks can occur due to causal confounding, with proofs for several combinatorial problems.

Contribution

It introduces a causal perspective to linear programming, demonstrating how confounder bias can lead to adversarial attacks, a novel approach in optimization problem analysis.

Findings

Confounder bias in LPs can be exploited for adversarial attacks.

Causal models explain influence of confounders on LP solutions.

Proofs provided for combinatorial and real-world problems.

Abstract

The recent years have been marked by extended research on adversarial attacks, especially on deep neural networks. With this work we intend on posing and investigating the question of whether the phenomenon might be more general in nature, that is, adversarial-style attacks outside classical classification tasks. Specifically, we investigate optimization problems as they constitute a fundamental part of modern AI research. To this end, we consider the base class of optimizers namely Linear Programs (LPs). On our initial attempt of a na\"ive mapping between the formalism of adversarial examples and LPs, we quickly identify the key ingredients missing for making sense of a reasonable notion of adversarial examples for LPs. Intriguingly, the formalism of Pearl's notion to causality allows for the right description of adversarial like examples for LPs. Characteristically, we show the direct influence of the Structural Causal Model (SCM) onto the subsequent LP optimization, which ultimately exposes a notion of confounding in LPs (inherited by said SCM) that allows for adversarial-style attacks. We provide both the general proof formally alongside existential proofs of such intriguing LP-parameterizations based on SCM for three combinatorial problems, namely Linear Assignment, Shortest Path and a real world problem of energy systems.