Constrained Training of Neural Networks via Theorem Proving.
Mark ChevallierMatthew WhyteJacques D. FleuriotPublished in: CoRR (2022)
Keyphrases
- theorem proving
- neural network
- training process
- first order logic
- automated reasoning
- training algorithm
- parallel search
- theorem prover
- decision procedures
- automated deduction
- term rewriting
- automatic programming
- proof planning
- inference rules
- automated theorem proving
- multi layer perceptron
- feedforward neural networks
- consequence finding
- proof procedure
- genetic algorithm
- back propagation
- distributed search
- geometry theorem proving
- combinatorial search
- logical inference
- supervised learning
- artificial neural networks
- training set
- search algorithm
- equational theories
- machine learning
- mathematical knowledge