Interpretable Apprenticeship Learning with Temporal Logic Specifications

Daniel Kasenberg and Matthias Scheutz

We consider the problem of inferring temporal logic specifications from agent behavior in stochastic domains. We formulate this as a multi-objective problem and use genetic programming to demonstrate the efficacy of our formluation.

@inproceedings{kasenbergscheutz17cdc,
  title={Interpretable Apprenticeship Learning with Temporal Logic Specifications},
  author={Daniel Kasenberg and Matthias Scheutz},
  year={2017},
  booktitle={Proceedings of the 56th IEEE Conference on Decision and Control (CDC 2017)},
  url={https://hrilab.tufts.edu/publications/kasenbergscheutz17cdc.pdf}
}