Computations and Equations for Segre-Grassmann Hypersurfaces

Basic Information

• Paper ID: 1408.2105
• Title: Computations and Equations for Segre-Grassmann Hypersurfaces
• Authors: Noah S. Daleo, Jonathan D. Hauenstein, Luke Oeding
• Classification: math.AG (Algebraic Geometry)
• Publication Date: August 19, 2015
• Paper Link: https://arxiv.org/abs/1408.2105

Abstract

In 2013, Abo and Wan studied the analogue of Waring's problem for systems of skew-symmetric forms and identified several defective systems. Of particular interest is when a certain secant variety of a Segre-Grassmann variety is expected to fill the natural ambient space, but is actually a hypersurface. Algorithms implemented in Bertini are used to determine the degrees of several of these hypersurfaces, and representation-theoretic descriptions of their equations are given. We answer Problem 6.5 Abo-Wan2013, and confirm their speculation that each member of an infinite family of hypersurfaces is minimally defined by a (known) determinantal equation. While led by numerical evidence, we provide non-numerical proofs for all of our results.

Research Background and Motivation

1. Core Problem: Investigating when secant varieties of Segre-Grassmann varieties are hypersurfaces and finding the defining equations of these hypersurfaces. This is related to the analogue of Waring's problem for systems of skew-symmetric forms.
2. Problem Significance:
   • Secant varieties are classical topics in algebraic geometry with broad applications in geometric complexity theory and signal processing
   • X-rank and X-border rank provide useful perspectives for tensors and related algebraic varieties
   • Describing the defining equations of σᵣ(X) allows easy determination of the X-border rank of any given point
3. Limitations of Existing Work:
   • Abo and Wan identified several defective systems in 2013, but left some open problems
   • In particular, Problem 6.5 remained unsolved, and conjectures regarding infinite families of hypersurfaces required verification
4. Research Motivation:
   • Combining numerical algebraic geometry and representation-theoretic tools to address these problems
   • Although research on certain skew-symmetric tensors is relatively limited, methods for finding border rank equations in one symmetric class often provide insights for another

Core Contributions

1. Resolved Abo-Wan's Problem 6.5: Proved that the prime ideal of the hypersurface σ₅(Seg(P²×G(2,5))) is generated by a single degree-6 polynomial
2. Confirmed Abo-Wan's Conjecture: Proved that each member of the infinite family of hypersurfaces σ₃ℓ₊₂(Seg(P²×G(1,4ℓ+2))) is minimally defined by a known determinantal equation
3. Computed Degrees of Multiple Hypersurfaces: Used Bertini algorithms to determine the exact degrees of several hypersurfaces
4. Provided Non-Numerical Proofs: Although guided by numerical evidence, rigorous algebraic proofs were provided for all results
5. Demonstrated the Power of Interdisciplinary Approaches: Successfully combined numerical algebraic geometry and representation-theoretic tools

Detailed Methodology

Problem Formulation

Study when the s-secant variety σₛ(Seg(Pᵐ×G(k,n))) of the Segre-Grassmann variety Seg(Pᵐ×G(k,n)) is a hypersurface, find its defining equations, and prove irreducibility.

Core Methodological Framework

1. Numerical Algebraic Geometry Approach

Using algorithms implemented in Bertini software to compute hypersurface degrees:

Hypersurface Degree Computation Algorithm:

• For an irreducible hypersurface H⊂Pⁿ, deg H = |H∩L|, where L is a general line
• Generate witness point sets using random monodromy loops
• Apply the trace test to verify completeness: each coordinate of ∑w∈W' pw(t) is a linear function of t

2. Representation-Theoretic Approach

Constructing invariant polynomials using Young symmetrizers:

Young Symmetrizer Algorithm:

• Begin with partitions (2,2,2) and (3,3,3,3,3,3)
• Find fillings of Young tableaux such that the corresponding Young symmetrizers produce nonzero images
• Construct polynomials through products of matrix determinants
• Replace variables using contraction operations

3. Exterior Flattening Construction

For infinite family cases, use Ottaviani-type exterior flattening:

• Construct a linear map φₜ: V⊗W* → V*⊗W for tensor T∈V⊗∧²W
• The map is antisymmetric in W and also antisymmetric in V
• det φₜ vanishes on the corresponding secant variety

Technical Innovations

1. Combining Numerical and Symbolic Methods: First use numerical methods to determine degrees, then construct exact equations using algebraic methods
2. Use of ⊠ Product: Distinguish from the usual Kronecker product ⊗, using ⊠ product to avoid the limitation det(P⊗Q) = det(P)ⁿdet(Q)ᵐ
3. Inductive Proof of Irreducibility: Prove irreducibility of determinants in large-scale cases through principal minor decomposition and induction

Experimental Setup

Computational Environment

• AMD Opteron 6376 processors (2.3 GHz cores)
• Single-core and multi-core parallel computation

Specific Cases

Investigation of the following specific hypersurfaces:

1. σ₅(Seg(P²×G(2,5))) ⊂ P⁵⁹
2. σ₅(Seg(P²×G(1,6))) ⊂ P⁶²
3. σ₈(Seg(P²×G(1,10))) ⊂ P¹⁶⁴
4. σ₁₁(Seg(P²×G(1,14))) ⊂ P³¹⁴

Verification Methods

• For small-scale cases: Direct computation in Macaulay2 using the factor command
• For medium-scale cases: Specialize variables to random lines and check factorization
• For large-scale cases: Use inductive proofs

Experimental Results

Main Computational Results

Numerical Results from Computation 3.1:

1. Degree of σ₅(Seg(P²×G(2,5))) ⊂ P⁵⁹ is 6
2. Degree of σ₅(Seg(P²×G(1,6))) ⊂ P⁶² is 21
3. Degree of σ₈(Seg(P²×G(1,10))) ⊂ P¹⁶⁴ is 33
4. Degree of σ₁₁(Seg(P²×G(1,14))) ⊂ P³¹⁴ is 45

Computational Efficiency

• First case: 6 random monodromy loops, 50 seconds computation time
• Subsequent three cases: 13, 12, and 13 loops respectively, with computation times of 2.5 minutes, 32 minutes, and 5.5 hours

Theoretical Results

Theorem 4.1: The prime ideal of the hypersurface σ₅(P²×G(2,5)) is generated by a single degree-6 polynomial constructed from specific Young symmetrizers, which has 10,080 monomials.

Theorem 5.3: For ℓ≥1, the prime ideal of the hypersurface σ₃ℓ₊₂(Seg(P²×G(1,4ℓ+2))) is generated by the determinant of a 3(4ℓ+3)×3(4ℓ+3) matrix φₜ.

Theorem 6.1: Regarding irreducibility of determinants of tensor products of generic matrices:

• For s=1,2: det(P⊠Q)=0
• For s=3: det(P⊠Q) factors as the cube of a cubic polynomial
• For s=4: det(P⊠Q) factors as the square of a degree-6 polynomial
• For s≥5: det(P⊠Q) is irreducible

Related Work

1. Classical Background: Alexander-Hirschowitz theorem resolves the case of Veronese embeddings
2. Segre Embeddings: Related problems have been studied in the case of Cartesian products, but remain incompletely resolved
3. Skew-Symmetric Versions: Related research in the skew-symmetric case, with some cases resolved
4. Equation Finding: Ottaviani's methods for symmetric tensors inspired new constructions for non-symmetric tensors

Conclusions and Discussion

Main Conclusions

1. Successfully resolved Abo-Wan's Problem 6.5
2. Confirmed conjectures regarding infinite families of hypersurfaces
3. Demonstrated the effectiveness of combining numerical algebraic geometry with representation theory

Limitations

1. For the equations of σ₅(Seg(P²×G(2,5))), no determinantal expression analogous to Ottaviani's degree-15 equation was found
2. Research on certain skew-symmetric tensors remains less developed than in the fully symmetric and non-symmetric cases

Future Directions

1. Find determinantal expressions for more hypersurfaces
2. Investigate possible connections with duals of Segre-Grassmann varieties
3. Apply the methods to problems in other symmetric classes

In-Depth Evaluation

Strengths

1. Methodological Innovation: Successfully combines numerical and symbolic computation, using numerical methods to guide and then providing rigorous proofs
2. Theoretical Depth: Involves representation theory, Young symmetrizers, invariant theory, and other deep mathematical tools
3. Computational Practicality: Provides concrete executable algorithms and software implementations
4. Completeness: Provides not only numerical evidence but also complete algebraic proofs

Weaknesses

1. Computational Complexity: Computation time grows rapidly for large-scale problems
2. Special Cases: Some small-scale cases require special treatment, lacking uniformity
3. Geometric Intuition: Lacks intuitive geometric interpretation of these hypersurfaces

Impact

1. Academic Value: Resolves important open problems in the field
2. Methodological Contribution: Demonstrates the power of interdisciplinary approaches, potentially inspiring solutions to other problems
3. Computational Tools: Provided algorithms and implementations have practical value for related research

Applicable Scenarios

1. Research on secant varieties in algebraic geometry
2. Tensor decomposition and rank problems
3. Geometric complexity theory
4. Algebraic problems requiring combined numerical and symbolic computation

References

The paper cites 46 related references, primarily including:

• Classical work by Alexander-Hirschowitz on Veronese embeddings
• Research by Abo and Wan on Waring's problem for skew-symmetric forms
• Work by Ottaviani on invariants and hyperdeterminants
• Related algorithms and software in numerical algebraic geometry