paper-with-me

홈 › Papers

Bounding the Complexity of Formally Verifying Neural Networks: A Geometric Approach

2020-12-22 · James Ferlez, Yasser Shoukry

In this paper, we consider the computational complexity of formally verifying the behavior of Rectified Linear Unit (ReLU) Neural Networks (NNs), where verification entails determining whether the NN satisfies convex polytopic specifications. Specifically, we show that for two different NN architectures -- shallow NNs and Two-Level Lattice (TLL) NNs -- the verification problem with (convex) polytopic constraints is polynomial in the number of neurons in the NN to be verified, when all other aspects of the verification problem held fixed. We achieve these complexity results by exhibiting explicit (but similar) verification algorithms for each type of architecture. Both algorithms efficiently translate the NN parameters into a partitioning of the NN's input space by means of hyperplanes; this has the effect of partitioning the original verification problem into polynomially many sub-verification problems derived from the geometry of the neurons. We show that these sub-problems may be chosen so that the NN is purely affine within each, and hence each sub-problem is solvable in polynomial time by means of a Linear Program (LP). Thus, a polynomial-time algorithm for the original verification problem can be obtained using known algorithms for enumerating the regions in a hyperplane arrangement. Finally, we adapt our proposed algorithms to the verification of dynamical systems, specifically when these NN architectures are used as state-feedback controllers for LTI systems. We further evaluate the viability of this approach numerically.

📄 PDF Abstract BibTeX arXiv:2012.11761

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Geoint-R1: Formalizing Multimodal Geometric Reasoning with Dynamic Auxiliary Constructions

2025-08-05 · Jingxuan Wei, Caijun Jia, Qi Chen, Honghao He 외 arxiv

Mathematical geometric reasoning is essential for scientific discovery and educational development, requiring precise logic and rigorous formal verification. While recent advances in Multimodal Large Language Models (MLL…

Multimodal Reasoning

ModelVerification.jl: a Comprehensive Toolbox for Formally Verifying Deep Neural Networks

2024-06-30 · Tianhao Wei, Luca Marzari, Kai S. Yun, Hanjiang Hu 외

Deep Neural Networks (DNN) are crucial in approximating nonlinear functions across diverse applications, ranging from image classification to control. Verifying specific input-output properties can be a highly challengin…

image-classificationImage Classification

Tight Bounds on $\ell_1$ Approximation and Learning of Self-Bounding Functions

2014-04-18 · Vitaly Feldman, Pravesh Kothari, Jan Vondrák

We study the complexity of learning and approximation of self-bounding functions over the uniform distribution on the Boolean hypercube ${0,1}^n$. Informally, a function $f:{0,1}^n \rightarrow \mathbb{R}$ is self-boundin…

Formally Verifying Analog Neural Networks Under Process Variations Using Polynomial Zonotopes

2026-05-11 · Yasmine Abu-Haeyeh, Tobias Ladner, Matthias Althoff, Lars Hedrich arxiv

Analog neural networks are gaining attention due to their efficiency in terms of power consumption and processing speed. However, since analog neural networks are implemented as physical circuits, they are highly sensiti…

Unaligned but Safe -- Formally Compensating Performance Limitations for Imprecise 2D Object Detection

2022-02-10 · Tobias Schuster, Emmanouil Seferis, Simon Burton, Chih-Hong Cheng

In this paper, we consider the imperfection within machine learning-based 2D object detection and its impact on safety. We address a special sub-type of performance limitations: the prediction bounding box cannot be perf…

2D Object Detectionobject-detectionObject Detection