John Clemens

Learning Device Models with Recurrent Neural Networks

2018

Recurrent neural networks (RNNs) are powerful constructs capable of modeling complex systems, up to and including Turing Machines. However, learning such complex models from finite training sets can be difficult. In this paper we empirically …

Runtime State Verification on Resource-Constrained Platforms

2018

Verifying the integrity of multi-component cyber and cyber-physical systems remains a crucial and necessary precondition for trust. The edge components of such systems, those that interact directly with the user or physical world, are typically …