Lower Bounds on the Homology of Vietoris–Rips Complexes of Hypercube Graphs

Henry Adams,Žiga Virk,Adams, Henry
DOI: https://doi.org/10.1007/s40840-024-01663-x
2024-03-05
Bulletin of the Malaysian Mathematical Sciences Society
Abstract:We provide novel lower bounds on the Betti numbers of Vietoris–Rips complexes of hypercube graphs of all dimensions and at all scales. In more detail, let be the vertex set of vertices in the n -dimensional hypercube graph, equipped with the shortest path metric. Let be its Vietoris–Rips complex at scale parameter , which has as its vertex set, and all subsets of diameter at most r as its simplices. For integers \(r<r'\) the inclusion is nullhomotopic, meaning no persistent homology bars have length longer than one, and we therefore focus attention on the individual spaces . We provide lower bounds on the ranks of homology groups of . For example, using cross-polytopal generators, we prove that the rank of is at least . We also prove a version of homology propagation : if and if p is the smallest integer for which , then for all . When , this result and variants thereof provide tight lower bounds on the rank of for all n , and for each we produce novel lower bounds on the ranks of homology groups. Furthermore, we show that for each , the homology groups of for contain propagated homology not induced by the initial cross-polytopal generators.</r'\)<>
mathematics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is about the lower bound estimation of the ranks of homology groups of Vietoris - Rips complexes of high - dimensional hypercube graphs at different scales. Specifically, the author provides new lower bounds for the Betti numbers of high - dimensional hypercube graphs under all scale parameters \(r\). This includes detailed estimations of the ranks of homology groups in specific dimensions, and deriving lower bounds of more extensive homology group ranks through the homology propagation technique. ### Main contributions of the paper 1. **New lower bounds of homology group ranks**: - The author proves that for all \(r\), \(\text{rank } H_{2r - 1}(\text{VR}(Q_n; r))\geq 2^n-(r + 1)\binom{n}{r+1}\). - Using cross - polytopal generators, the author proves the above inequality. 2. **Homology propagation technique**: - The author introduces a method called "homology propagation", through which the homology information of low - dimensional hypercube graphs can be extended to higher - dimensional cases. - Specifically, if \(p\) is the smallest integer such that \(\text{rank } H_q(\text{VR}(Q_p; r))\neq 0\), then for all \(n\geq p\), we have \[ \text{rank } H_q(\text{VR}(Q_n; r))\geq\sum_{i = p}^n2^{i - p}\binom{i - 1}{p - 1}\cdot\text{rank } H_q(\text{VR}(Q_p; r)). \] 3. **Results at specific scales**: - When \(r\leq3\), the author's results provide tight lower bounds for \(\text{rank } H_q(\text{VR}(Q_n; r))\) for all \(n\). - For \(r\geq4\), the author provides new lower bounds of homology group ranks. 4. **Study of homotopy types**: - The author discusses the homotopy types of \(\text{VR}(Q_n; r)\), especially the known results when \(r\leq3\). - For example, when \(r = 0\), \(\text{VR}(Q_n; 0)\) is the disjoint union of \(2^n\) vertices and is homotopy equivalent to the wedge sum of \((2^n - 1)\) zero - dimensional spheres. - When \(r = 1\), \(\text{VR}(Q_n; 1)\) is a connected graph and is homotopy equivalent to the wedge sum of \(((n - 2)2^{n - 1}+1)\) circles. - When \(r = 2\), \(\text{VR}(Q_n; 2)\) is homotopy equivalent to the wedge sum of several three - dimensional spheres. - When \(r = 3\), \(\text{VR}(Q_n; 3)\) is homotopy equivalent to the wedge sum of several seven - dimensional spheres and four - dimensional spheres. ### Conclusion This paper provides new lower bound estimations for the ranks of homology groups of Vietoris - Rips complexes of high - dimensional hypercube graphs at different scales by introducing new techniques and methods. These results not only fill the gaps in existing research but also provide a theoretical basis for further research in the future.