Session 8. Machine Learning for Mathematical Software


Aim and Scope

While there has been some discussion on how Computer Algebra Systems could be used for AI there is little literature on applications in the other direction. However, recent results for quantifier elimination suggest that given enough example problems there is scope for tools like Support Vector Machines to improve the performance of Computer Algebra Systems. It may seem that the inherently probabilistic nature of machine learning tools would invalidate the exact mathematical results prized by such software, however, algorithms and implementations often come with a range of choices which have no effect on the mathematical correctness of the end result but a great effect on the resources required to find it. This sessions aims to promote thought on whether, where, and how best machine learning could be applied to mathematical software development.

Accepted Talks

Erika Abraham: Heuristics in SMT Solving: To Learn or not to Learn?

Yihe Dong: NLP-based Detection of Mathematical Subject Classification

Matthew England: Machine Learning for Mathematical Software

Munehiro Kobayashi, Hidenao Iwane, Takuya Matsuzaki and Hirokazu Anai: Ordering of Subformulas for Efficient Quantifier Elimination over Real Closed Field

Jonathan Gryak, Robert Haralick and Delaram Kahrobaei: Solving Algorithmic Problems in Algebraic Structures via Machine Learning

Thomas Sturm: Do Reduce Switches Point at Suitable Spots for Machine Learning?

Stephen A. Forrest: Integration of Deep Learning in Maple

Josef Urban: A Brief Overview of Machine Learning for Automated Reasoning.