Forward and inverse design of kirigami via supervised autoencoder
Paul Z. Hanakata, Ekin D. Cubuk, David Campbell, Harold S. Park
Abstract
The authors develop a supervised autoencoder to perform forward and inverse design of graphene kirigami. By interpolating in the latent space of kirigami structures, the supervised autoencoder generates designs that mix parallel and orthogonal cuts.
Topics & Concepts
AutoencoderInverseComputer scienceArtificial intelligenceMathematicsArtificial neural networkGeometryAdvanced Materials and MechanicsTeaching and Learning Programming