Output Reachable Set Estimation and Verification for Multilayer Neural Networks
Top Cited Papers
- 15 March 2018
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Networks and Learning Systems
- Vol. 29 (11) , 5777-5783
Abstract
In this brief, the output reachable estimation and safety verification problems for multilayer perceptron (MLP) neural networks are addressed. First, a conception called maximum sensitivity is introduced, and for a class of MLPs whose activation functions are monotonic functions, the maximum sensitivity can be computed via solving convex optimization problems. Then, using a simulation-based method, the output reachable set estimation problem for neural networks is formulated into a chain of optimization problems. Finally, an automated safety verification is developed based on the output reachable set estimation result. An application to the safety verification for a robotic arm model with two joints is presented to show the effectiveness of the proposed approaches.Keywords
All Related Versions
Funding Information
- National Science Foundation (CNS 1464311, CNS 1713253, SHF 1527398, SHF 1736323)
- Air Force Office of Scientific Research (FA9550-15-1-0258, FA9550-16-1-0246, FA9550-18-1-0122)
This publication has 26 references indexed in Scilit:
- A non-ellipsoidal reachable set estimation for uncertain neural networks with time-varying delayCommunications in Nonlinear Science and Numerical Simulation, 2014
- Architecture selection for networks trained with extreme learning machine using localized generalization error modelNeurocomputing, 2013
- An Abstraction-Refinement Approach to Verification of Artificial Neural Networks2010
- Sensitivity analysis applied to the construction of radial basis function networksNeural Networks, 2005
- A Quantified Sensitivity Measure for Multilayer Perceptron to Input PerturbationNeural Computation, 2003
- Sensitivity analysis of multilayer perceptron to input and weight perturbationsIEEE Transactions on Neural Networks, 2001
- Adaptive neural network control of nonlinear systems by state and output feedbackIEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 1999
- Face recognition: a convolutional neural-network approachIEEE Transactions on Neural Networks, 1997
- Neural networks for control systems—A surveyAutomatica, 1992
- Multilayer feedforward networks are universal approximatorsNeural Networks, 1989