2112.12295
From Finite Vector Field Data to Combinatorial Dynamical Systems in the Sense of Forman
Dominic Desjardins Côté
correctmedium confidence
- Category
- Not specified
- Journal tier
- Specialist/Solid
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper defines the ILP and cost matrix and claims that any global minimizer induces a combinatorial dynamical system (Theorem 1.1/5.1) . Its proof argues (i) the constraint Σ_i a_{ik} + Σ_j a_{kj} − a_{kk} = 1 enforces that each simplex participates exactly once, yielding Dom V ∪ Im V = K and Dom V ∩ Im V = Crit V, and (ii) any non-admissible off-diagonal a_{ij}=1 can be replaced by a_{ii}=a_{jj}=1 at strictly lower cost because C_{ij} = max(2α+1,3) > 2α = C_{ii}+C_{jj} . The model’s solution gives the same two steps, but more cleanly via row/column sums (r_k, c_k) to establish partial injectivity and the Dom/Im identities, and then the local replacement argument for non-admissible pairs. Minor wording imprecision in the paper (“admissible matching” vs. “matching feasibility”) aside, both arguments establish the same result with essentially the same proof skeleton.
Referee report (LaTeX)
\textbf{Recommendation:} minor revisions \textbf{Journal Tier:} specialist/solid \textbf{Justification:} Mathematically sound and implementable. The proof that an optimal ILP solution yields a combinatorial dynamical system follows directly from the nodewise equality constraint and the cost design that penalizes non-admissible pairs. Exposition can be improved by explicitly formalizing the row/column-sum argument and clearly separating feasibility (matching) from admissibility (objective) reasoning. With these clarifications, the paper serves as a solid reference for constructing Forman-style dynamics from data.