Projects per year
This chapter argues for a novel method to machine learn patterns in formal proofs using statistical machine learning methods. The method exploits coalgebraic approach to proofs. The success of the method is demonstrated on three applications allowing to distinguish well-formed proofs from ill-formed proofs, identify families of proofs and even families of potentially provable goals.
Coalgebraic Logic Programming for Type Inference: Parallelism and Corecursion for New Generation of Programming Languages (Joint with the University of Bath)
1/09/13 → 31/01/17