Overview
This research addresses the problem of minimal parametric generalized fillings within finite metric spaces. This problem is framed as a version of the optimal connection problem. It establishes a connection between these fillings and the construction of convex multidimensional polyhedra. Specifically, for each tree $(G)$ connecting a finite metric space $(M)$, a corresponding convex multidimensional polyhedron $(W_G)$ is constructed. The weight associated with a parametric minimal filling is determined by identifying the maximum value of a specific linear function on the polyhedron $(W_G)$. This linear function corresponds to the distance vector of the metric space $(M)$. A key finding demonstrates that the union of the vertex sets of all possible polyhedra $(W_G)$, formed over all possible choices of $(G)$, constitutes an extremal subset. This implies that this union precisely matches the vertex set of its own convex hull.
Research Context
The study operates within the theoretical domain of finite metric spaces and combinatorial optimization, particularly focusing on optimal connection problems. The concept of "parametric generalized fillings" extends traditional notions of connecting points within a metric space, introducing a parameter that influences the filling's properties. These fillings are conceptualized through their connection to geometric structures, specifically convex polyhedra. The structure $(G)$ is described as a tree type chosen for the parametric filling, indicating that the filling process involves graph-theoretic elements where connectivity is represented by a tree. The distance vector of $M$ serves as a critical input, reflecting the inherent distances between points in the metric space, which in turn influences the linear function used for evaluating filling weights. Understanding the relationship between the graph structure $(G)$, the resulting polyhedra $(W_G)$, and the properties of minimal fillings is central to this context.
Approach
The approach involves constructing a convex multidimensional polyhedron $(W_G)$ for each tree $(G)$ that connects the finite metric space $(M)$. This construction is directly linked to the problem of minimal parametric generalized fillings. The weight of a parametric minimal filling is then derived. This derivation is accomplished by computing the maximum of a special linear function over the constructed polyhedron $(W_G)$. This linear function incorporates the distance vector of $(M)$. The investigation proceeds by considering the collection of all possible trees $(G)$ that can connect $(M)$, and subsequently, the corresponding set of polyhedra $(W_G)$. A key proof within the research concerns the properties of the vertex sets of these polyhedra. The methodology establishes that the combined set of vertices from all $(W_G)$ forms an extremal set, meaning it coincides with the vertex set of its own convex hull. This step involves analyzing the geometric properties and relationships between these polyhedra as defined by different tree structures $(G)$.
Findings
- For each tree $(G)$ connecting a finite metric space $(M)$, a convex multidimensional polyhedron $(W_G)$ is constructible.
- The weight of a parametric minimal filling can be determined as the maximum value of a specific linear function on $(W_G)$. This linear function corresponds to the distance vector of $(M)$.
- The union of the vertex sets of the polyhedra $(W_G)$, aggregated over all possible trees $(G)$, forms an extremal subset.
- This extremal subset demonstrably coincides with the vertex set of its own convex hull.