Neural Network Branch-and-Bound for Neural Network Verification

07/27/2021
by   Florian Jaeckle, et al.
0

Many available formal verification methods have been shown to be instances of a unified Branch-and-Bound (BaB) formulation. We propose a novel machine learning framework that can be used for designing an effective branching strategy as well as for computing better lower bounds. Specifically, we learn two graph neural networks (GNN) that both directly treat the network we want to verify as a graph input and perform forward-backward passes through the GNN layers. We use one GNN to simulate the strong branching heuristic behaviour and another to compute a feasible dual solution of the convex relaxation, thereby providing a valid lower bound. We provide a new verification dataset that is more challenging than those used in the literature, thereby providing an effective alternative for testing algorithmic improvements for verification. Whilst using just one of the GNNs leads to a reduction in verification time, we get optimal performance when combining the two GNN approaches. Our combined framework achieves a 50% reduction in both the number of branches and the time required for verification on various convolutional networks when compared to several state-of-the-art verification methods. In addition, we show that our GNN models generalize well to harder properties on larger unseen networks.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/03/2019

Neural Network Branching for Neural Network Verification

Formal verification of neural networks is essential for their deployment...
research
01/27/2022

Algorithm Selection for Software Verification using Graph Attention Networks

The field of software verification has produced a wide array of algorith...
research
06/10/2022

We Cannot Guarantee Safety: The Undecidability of Graph Neural Network Verification

Graph Neural Networks (GNN) are commonly used for two tasks: (whole) gra...
research
03/07/2022

Scalable Verification of GNN-based Job Schedulers

Recently, Graph Neural Networks (GNNs) have been applied for scheduling ...
research
04/30/2022

Complete Verification via Multi-Neuron Relaxation Guided Branch-and-Bound

State-of-the-art neural network verifiers are fundamentally based on one...
research
06/30/2022

Lookback for Learning to Branch

The expressive and computationally inexpensive bipartite Graph Neural Ne...
research
02/24/2020

Lagrangian Decomposition for Neural Network Verification

A fundamental component of neural network verification is the computatio...

Please sign up or login with your details

Forgot password? Click here to reset