Thoughts on Generative Grammars and Their Use in Automated Theorem Proving Based on Neural Networks (With a Type-0 Grammar of Nicod’s Propositional Logic System Added)
Elmar Guseinov · viXra · 2021
The article describes a new method of applying neural networks in theorem proving. The main idea is to use the generative grammars of first-order theories. Thus, basic theories (e.g., group or number theory) can be concerned. The corresponding program generates as much training data as one may need in machine learning. All the necessary information on logic, grammars, and neural networks was given. Also, the grammar of propositional logic was first described.