[Paper Review] Generalizations of Joints Problem

Ben Yang|arXiv (Cornell University)|Jun 28, 2016

Computational Geometry and Mesh Generation15 references3 citations

TL;DR

This paper generalizes the joints problem to higher-dimensional algebraic varieties, proving an almost sharp bound on the number of joints formed by sets of $\alpha_i$-dimensional varieties in $\mathbb{R}^n$, where $\sum \alpha_i = n$. Using polynomial partitioning and induction, it establishes that the number of joints is $O((\prod |\mathcal{S}_i|)^{1/(k-1) + \epsilon})$ for $k$ families of varieties, with a special case showing $O(N^{3/2 + \epsilon})$ joints for $N$ 2-planes in $\mathbb{R}^6$. The result extends the classical joints theorem to higher-dimensional objects and provides a framework for counting multijoints with multiplicities.

ABSTRACT

We generalize the joints problem to sets of varieties and prove almost sharp bound on the number of joints. As a special case, given a set of $N$ $2$-planes in $\mathbb{R}^6$, the number of points at which three $2$-planes intersect and span $\mathbb{R}^6$ is at most $CN^{3/2+ε}$. We also get almost sharp bound on the number of joints with multiplicities. The main tools are polynomial partitioning and induction on dimension.

Motivation & Objective

• To extend the classical joints problem—originally about lines in R^3—to higher-dimensional algebraic varieties, such as 2-planes in R^6.
• To establish an almost sharp upper bound on the number of joints formed by multiple families of varieties in R^n, where each variety has dimension α_i and ∑α_i = n.
• To generalize the notion of joints to multijoints formed by transversal families of varieties and to bound the number of such joints with multiplicities.
• To address the challenge that higher-dimensional objects (e.g., planes) can intersect a polynomial zero set infinitely, requiring new techniques beyond line-based methods.

Proposed method

• Uses polynomial partitioning to divide R^n into cells, each intersecting only a bounded number of varieties from each family, enabling inductive control over joints in each cell.
• Applies induction on dimension and the number of families, reducing the problem in n dimensions to lower-dimensional subproblems via projection and transversality.
• Employs Hölder’s inequality and bounds on line intersections with algebraic varieties to control the sum of joint multiplicities over smooth points of the zero set of the partitioning polynomial.
• Decomposes the zero set of the partitioning polynomial into lower-dimensional smooth varieties using a result from algebraic geometry, allowing recursive application of the induction hypothesis.
• Imposes transversality conditions ensuring that tangent spaces of varieties at a joint span R^n, which is essential for the inductive argument and for avoiding degenerate configurations.
• Introduces a weighted sum over joints involving the product of multiplicities $\prod N_i(x)^{1/(n-1)}$, which serves as a discrete analogue of multilinear Kakeya estimates.

Experimental results

Research questions

• RQ1 What is the maximum number of joints that can be formed by k families of α_i-dimensional varieties in R^n, where ∑α_i = n and the varieties are transversal at each joint?
• RQ2 Can the classical joints theorem for lines in R^3 be generalized to higher-dimensional objects such as 2-planes in R^6?
• RQ3 How do joint multiplicities behave in higher-dimensional settings, and can they be bounded in a way analogous to the multilinear Kakeya inequality?
• RQ4 What modifications are needed in the polynomial method when the objects are not 1-dimensional curves or lines, but higher-dimensional algebraic varieties?
• RQ5 Is it possible to achieve an almost sharp bound of the form $O((\prod |\mathcal{S}_i|)^{1/(k-1) + \epsilon})$ for joints formed by families of varieties of bounded degree and dimension?

Key findings

• The number of joints formed by k families of α_i-dimensional varieties in R^n, with ∑α_i = n and each variety defined by bounded-degree polynomials, is bounded by $C(n,m,d,\epsilon)(\prod |\mathcal{S}_i|)^{1/(k-1) + \epsilon}$ for any ε > 0.
• For 2-planes in R^6, the number of joints formed by three families of N 2-planes each is $O(N^{3/2 + \epsilon})$, matching the expected order up to an ε-loss.
• The bound is almost sharp, as constructions based on axis-parallel configurations in lower dimensions yield matching lower bounds up to ε.
• The paper proves an almost sharp bound on the sum of joint multiplicities, specifically $\sum_x \left(\prod_{i=1}^n N_i(x)\right)^{1/(n-1)} = O\left(\left(\prod L_i\right)^{1/(n-1) + \epsilon}\right)$, extending a conjecture by Carbery to higher dimensions.
• The method successfully handles the challenge that higher-dimensional varieties can intersect a polynomial zero set infinitely, by decomposing the zero set into lower-dimensional smooth components and applying induction.
• The result generalizes previous work on joints for lines and curves to arbitrary varieties, and provides a framework for future extensions to fields other than R.