Show simple item record

dc.contributor.authorFan, Jiamengen_US
dc.contributor.authorHuang, Chaoen_US
dc.contributor.authorLi, Wenchaoen_US
dc.contributor.authorChen, Xinen_US
dc.contributor.authorZhu, Qien_US
dc.date.accessioned2020-05-19T13:14:46Z
dc.date.available2020-05-19T13:14:46Z
dc.date.issued2019-11
dc.identifier.citationJiameng Fan, Chao Huang, Wenchao Li, Xin Chen, Qi Zhu. 2019. "Towards Verification-Aware Knowledge Distillation for Neural-Network Controlled Systems: Invited Paper." 2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD). 2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD). 2019-11-04 - 2019-11-07. https://doi.org/10.1109/iccad45719.2019.8942059
dc.identifier.urihttps://hdl.handle.net/2144/41005
dc.description.abstractNeural networks are widely used in many applications ranging from classification to control. While these networks are composed of simple arithmetic operations, they are challenging to formally verify for properties such as reachability due to the presence of nonlinear activation functions. In this paper, we make the observation that Lipschitz continuity of a neural network not only can play a major role in the construction of reachable sets for neural-network controlled systems but also can be systematically controlled during training of the neural network. We build on this observation to develop a novel verification-aware knowledge distillation framework that transfers the knowledge of a trained network to a new and easier-to-verify network. Experimental results show that our method can substantially improve reachability analysis of neural-network controlled systems for several state-of-the-art toolsen_US
dc.language.isoen_US
dc.publisherIEEEen_US
dc.relation.ispartof2019 IEEE/ACM International Conference on Computer-Aided Design (ICCAD)
dc.subjectFormal verificationen_US
dc.subjectNeurocontrollersen_US
dc.subjectReachability analysisen_US
dc.subjectTrained networken_US
dc.subjectLipschitz continuityen_US
dc.titleTowards verification-aware knowledge distillation for neural-network controlled systems: invited paperen_US
dc.typeConference materialsen_US
dc.description.versionAccepted manuscripten_US
dc.identifier.doi10.1109/iccad45719.2019.8942059
pubs.elements-sourcecrossrefen_US
pubs.notesEmbargo: Not knownen_US
pubs.organisational-groupBoston Universityen_US
pubs.organisational-groupBoston University, College of Engineeringen_US
pubs.organisational-groupBoston University, College of Engineering, Department of Electrical & Computer Engineeringen_US
pubs.publication-statusPublisheden_US
dc.identifier.orcid0000-0003-0153-4648 (Li, Wenchao)
dc.identifier.mycv550021


This item appears in the following Collection(s)

Show simple item record