Using Machine Learning to Decide When to Precondition Cylindrical Algebraic Decomposition with Groebner Bases

@article{Huang2016UsingML,
  title={Using Machine Learning to Decide When to Precondition Cylindrical Algebraic Decomposition with Groebner Bases},
  author={Zongyan Huang and Matthew England and James H. Davenport and Lawrence Charles Paulson},
  journal={2016 18th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC)},
  year={2016},
  pages={45-52}
}
Cylindrical Algebraic Decomposition (CAD) is a key tool in computational algebraic geometry, particularly for quantifier elimination over real-closed fields. However, it can be expensive, with worst case complexity doubly exponential in the size of the input. Hence it is important to formulate the problem in the best manner for the CAD algorithm. One possibility is to precondition the input polynomials using Groebner Basis (GB) theory. Previous experiments have shown that while this can often… 

Figures and Tables from this paper

Using Machine Learning to Improve Cylindrical Algebraic Decomposition

This paper applies machine learning to Symbolic Computation in two case studies: the first to select between heuristics for choosing a CAD variable ordering; the second to identify when a CAD problem instance would benefit from Gröbner Basis preconditioning.

Algorithmically generating new algebraic features of polynomial systems for machine learning

This work created a framework to describe all the previously identified ML features for the problem and enumerated all options in this framework to automatically generation many more features, demonstrating the usefulness of these with an experiment which shows that an ML choice for CAD variable ordering is superior to those made by human created heuristics.

Comparing machine learning models to choose the variable ordering for cylindrical algebraic decomposition

The problem of selecting the variable ordering for cylindrical algebraic decomposition (CAD), an important algorithm in Symbolic Computation, is addressed by training ML models to select theVariable ordering directly, and by trying out a wider variety of ML techniques.

The Complexity of Cylindrical Algebraic Decomposition with Respect to Polynomial Degree

Cylindrical algebraic decomposition (CAD) is an important tool for working with polynomial systems, particularly quantifier elimination. However, it has complexity doubly exponential in the number of

Applying Machine Learning to Heuristics for Real Polynomial Constraint Solving

This paper considers the application of machine learning to automatically generating heuristics for real polynomial constraint solvers. We consider a specific choice-point in the algorithm for

Machine-Learned Preconditioners for Linear Solvers in Geophysical Fluid Flows

It is tested whether machine learning methods can be used for preconditioning to increase the performance of the linear solver -- the backbone of the semi-implicit, grid-point model approach for

Improved cross-validation for classifiers that make algorithmic choices to minimise runtime without compromising output correctness

A modification to the cross-validation algorithms of the classifiers are investigated so that they do distinguish between any of the non-optimal orderings, leading to improved results.

Machine Learning for Mathematical Software

The authors own work and similar applications for other mathematical software are surveyed, suggesting that, given enough example problems, there is scope for machine learning tools like Support Vector Machines to improve the performance of Computer Algebra Systems.

Experience with Heuristics, Benchmarks & Standards for Cylindrical Algebraic Decomposition

This work summarises the experience with heuristic development for the computer algebra algorithm Cylindrical Algebraic Decomposition and proposes and discusses standards and benchmarks as another area where Symbolic Computation could prosper from Satisfiability Checking expertise.

References

SHOWING 1-10 OF 66 REFERENCES

Applying machine learning to the problem of choosing a heuristic to select the variable ordering for cylindrical algebraic decomposition

This paper uses machine learning (specifically a support vector machine) to select between heuristics for choosing a variable ordering, outperforming each of the separateHeuristics.

A repository for CAD examples

A unified machine-readable repository of examples to be used when considering CADs, which allows for quick access to a host of examples, pulled from various theoretical questions and applications.

Optimising Problem Formulation for Cylindrical Algebraic Decomposition

A new measure for CAD complexity is suggested which takes into account the real geometry of the problem and leads to new heuristics for choosing: the variable ordering for a CAD problem, a designated equational constraint, and formulations for truth-table invariant CADs (TTICADs).

Optimising Problem Formulation for Cylindrical Algebraic Decomposition

Cylindrical algebraic decomposition (CAD) is an important tool for the study of real algebraic geometry with many applications both within mathematics and elsewhere. It is known to have doubly

Truth table invariant cylindrical algebraic decomposition

Speeding Up Cylindrical Algebraic Decomposition by Gröbner Bases

It is seen that it is possible to apply Grobner bases to the (conjoined) equalities before invoking CAD, and this is, quite often, a beneficial preconditioning of the CAD problem.

Choosing a Variable Ordering for Truth-Table Invariant Cylindrical Algebraic Decomposition by Incremental Triangular Decomposition

The latest variant of this which builds the complex decomposition incrementally by polynomial and produces CADs on whose cells a sequence of formulae are truth-invariant is considered.

The Complexity of Cylindrical Algebraic Decomposition with Respect to Polynomial Degree

Cylindrical algebraic decomposition (CAD) is an important tool for working with polynomial systems, particularly quantifier elimination. However, it has complexity doubly exponential in the number of

Cylindrical algebraic decompositions for boolean combinations

This article generalises the theory of equational constraints to design an algorithm which will efficiently construct a TTICAD for a wide class of problems, producing stronger results than when using equational constraint alone.

Constructing fewer open cells by GCD computation in CAD projection

It is proved that the new projection operator based on cylindrical algebraic decomposition can be used for testing semi-definiteness of polynomials and still guarantees obtaining at least one sample point from every connected component of the highest dimension.
...