Machine Learning of Coq Proof Guidance: First Experiments

8 pages•Published: December 4, 2014

Cezary Kaliszyk, Lionel Mamane and Josef Urban

Abstract

We report the results of the first experiments with learning proof dependencies from the formalizations done with the Coq system. We explain the process of obtaining the dependencies from the Coq proofs, the characterization of formulas that is used for the learning, and the evaluation method. Various machine learning methods are compared on a dataset of 5021 toplevel Coq proofs coming from the CoRN repository. The best resulting method covers on average 75% of the needed proof dependencies among the first 100 predictions, which is a comparable performance of such initial experiments on other large-theory corpora.

Keyphrases: coq, formal mathematics, interactive theorem proving, machine learning, proof advice, type theory

In: Temur Kutsia and Andrei Voronkov (editors). SCSS 2014. 6th International Symposium on Symbolic Computation in Software Science, vol 30, pages 27-34.

Links:	https://easychair.org/publications/paper/2gK
	https://doi.org/10.29007/lmmg

BibTeX entry

@inproceedings{SCSS2014:Machine_Learning_Coq_Proof,
  author    = {Cezary Kaliszyk and Lionel Mamane and Josef Urban},
  title     = {Machine Learning of Coq Proof Guidance: First Experiments  },
  booktitle = {SCSS 2014. 6th International Symposium on Symbolic Computation in Software Science},
  editor    = {Temur Kutsia and Andrei Voronkov},
  series    = {EPiC Series in Computing},
  volume    = {30},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2398-7340},
  url       = {/publications/paper/2gK},
  doi       = {10.29007/lmmg},
  pages     = {27-34},
  year      = {2014}}

Download PDF Open PDF in browser