The Differential Counting Polynomial

Markus Lange-HegermannFound. Comput. Math 2018 · OpenReview

TL;DR

This paper introduces the "differential counting polynomial," a new tool for quantitatively analyzing the solution sets of systems of differential equations. This polynomial extends existing algebraic tools and can help determine if a found set of solutions is the complete set, a crucial problem when symbolic solutions are hard to obtain.

What Problem Does This Paper Address?

Many systems of differential equations, especially nonlinear ones, are notoriously difficult to solve analytically. Computer algebra systems can often find some solutions, but a fundamental question remains: are these all the solutions? This paper tackles the challenge of rigorously verifying the completeness of solution sets for both ordinary and partial differential equations. Existing methods, like Kolchin's dimension polynomial, provide a measure of solution set size but lack the granularity to distinguish between subtly different solution structures, such as countable infinite exceptional sets. The authors aim to develop a more refined tool to address this.

Methods

The core contribution is the introduction of the differential counting polynomial. This concept builds upon Plesken's algebraic counting polynomial, which measures the size of constructible sets in algebraic geometry. The differential counting polynomial is defined by extending the algebraic counting polynomial's axioms to incorporate differential equations.

Key methodological steps include:

  • Generalizing Algebraic Counting Polynomials: The paper first recaps and extends the algebraic counting polynomial to handle constructible sets that might include countably infinite complements.
  • Defining the Differential Counting Polynomial: This involves defining an "algebraically restricted $\sigma$-system of differential equations" and then applying a process analogous to the algebraic Thomas decomposition to formal power series solutions. This leads to the concept of a "counting sequence" and, under certain conditions, a "differential counting polynomial."
  • Connecting to Differential Dimension Polynomial: The paper establishes relationships between the differential counting polynomial and the well-established differential dimension polynomial, showing how the former provides finer details.
  • Developing Decision Procedures: Theorems are presented that use the differential counting polynomial to decide equality of solution sets, particularly for "well-fibered" sets.

Datasets

Not discussed in the paper. The paper focuses on theoretical development and provides examples using constructed systems of differential equations rather than empirical datasets.

Key Insights and Ideas

  • Beyond Dimension: The differential counting polynomial offers a more detailed "measure" of solution sets than the differential dimension polynomial, capable of capturing finer structures like countable infinite exceptional sets.
  • Verifying Solution Completeness: The primary motivation and insight is that the differential counting polynomial can be used to determine if a given set of solutions is indeed the complete set of solutions for a system of differential equations.
  • Generalization of Algebraic Tools: The work successfully bridges algebraic geometry tools (counting polynomials) with the theory of differential equations, creating a powerful new analytical instrument.
  • The Role of $\aleph_0$: The indeterminate $\aleph_0$ (representing countable infinity) plays a crucial role in the differential counting polynomial, directly reflecting the presence of countable exceptional sets in solutions.
  • Example-Driven Development: The paper uses detailed examples, particularly the Navier-Stokes equations and a specific ODE system, to illustrate the concepts and demonstrate the practical implications of the differential counting polynomial.

Key Results

  • Definition of Differential Counting Polynomial: A formal definition and framework for the differential counting polynomial are established.
  • Decision Criteria for Set Equality: Theorems (e.g., Theorem 4.3) are proven showing that under certain conditions, equality of differential counting polynomials implies equality of solution sets.
  • Relationship to Differential Dimension Polynomial: Theorem 4.5 shows that the differential counting polynomial generalizes the differential dimension polynomial, with its degree and leading coefficients relating to differential type, typical dimension, and differential dimension.
  • Formula for Simple Systems: Theorem 4.6 provides a closed-form formula for the differential counting polynomial of simple differential systems without inequations.
  • Illustrative Examples: The paper demonstrates through examples (like Example 4.8 and 4.9) how the differential counting polynomial can reveal that symbolic solvers might miss certain solutions, even when the dimension polynomial suggests otherwise.

Limitations and Future Work

  • Existence is Not Guaranteed: The paper explicitly states that the existence of a differential counting sequence or polynomial is not clear in general. This is a significant limitation, as the tool is not universally applicable.
  • Algorithmic Determination: Determining the differential counting polynomial is not algorithmic in general, even for simpler cases like single linear differential equations, where it can be reduced to Hilbert's tenth problem.
  • Computational Complexity: While the paper defines the polynomial, its practical computation for complex systems is likely to be challenging.
  • Future Work: The paper suggests that further exploration of (classes of) examples is needed. The authors' PhD thesis [14] is mentioned as containing additional examples, hinting at ongoing research in this area. The proofs of several theorems are postponed to later sections or appendices, indicating the depth and complexity of the underlying mathematics.

Summary updated

Abstract

The aim of this paper is a quantitative analysis of the solution set of a system of polynomial nonlinear differential equations, both in the ordinary and partial case. Therefore, we introduce the differential counting polynomial, a common generalization of the dimension polynomial and the (algebraic) counting polynomial. Under mild additional assumptions, the differential counting polynomial decides whether a given set of solutions of a system of differential equations is the complete set of solutions.
BibTeX
@inproceedings{langehegermann2018differential,
  author = {Markus Lange-Hegermann},
  title = {{The Differential Counting Polynomial}},
  booktitle = {Found. Comput. Math},
  year = {2018},
  url = {https://openreview.net/forum?id=N0eYSYlBcTq},
}
Figure