Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Lagrangian Decomposition for Neural Network Verification
(Submitted on 24 Feb 2020 (v1), last revised 17 Jun 2020 (this version, v3))
Abstract: A fundamental component of neural network verification is the computation of bounds on the values their outputs can take. Previous methods have either used off-the-shelf solvers, discarding the problem structure, or relaxed the problem even further, making the bounds unnecessarily loose. We propose a novel approach based on Lagrangian Decomposition. Our formulation admits an efficient supergradient ascent algorithm, as well as an improved proximal algorithm. Both the algorithms offer three advantages: (i) they yield bounds that are provably at least as tight as previous dual algorithms relying on Lagrangian relaxations; (ii) they are based on operations analogous to forward/backward pass of neural networks layers and are therefore easily parallelizable, amenable to GPU implementation and able to take advantage of the convolutional structure of problems; and (iii) they allow for anytime stopping while still providing valid bounds. Empirically, we show that we obtain bounds comparable with off-the-shelf solvers in a fraction of their running time, and obtain tighter bounds in the same time as previous dual algorithms. This results in an overall speed-up when employing the bounds for formal verification. Code for our algorithms is available at this https URL
Submission history
From: Alessandro De Palma [view email][v1] Mon, 24 Feb 2020 17:55:10 GMT (2659kb,D)
[v2] Sun, 5 Apr 2020 18:00:02 GMT (1741kb,D)
[v3] Wed, 17 Jun 2020 17:49:58 GMT (7201kb,D)
Link back to: arXiv, form interface, contact.