Constrained Training of Neural Networks via Theorem Proving

no code implementations • 8 Jul 2022 • Mark Chevallier, Matthew Whyte, Jacques D. Fleuriot

We introduce a theorem proving approach to the specification and generation of temporal logical constraints for training neural networks.

Automated Theorem Proving Code Generation

Formalising the Foundations of Discrete Reinforcement Learning in Isabelle/HOL

no code implementations • 11 Dec 2021 • Mark Chevallier, Jacques Fleuriot

We present a formalisation of finite Markov decision processes with rewards in the Isabelle theorem prover.

reinforcement-learning Reinforcement Learning (RL)