Algorithms for Standard-form ILP Problems via Komlós' Discrepancy Setting (Refined $2^{O(k)}$-analysis)
We study the standard-form ILP problem $\max\{ c^\top x \colon A x = b,\; x \in Z_{\geq 0}^n \}$, where $A\in Z^{k\times n}$ has full row rank. We obtain refined FPT algorithms parameterized by $k$ and $Δ$, the maximum absolute value of a $k\times k$ minor of $A$. Our approach combines discrepancy-based dynamic programming with matrix $\ell_2$-discrepancy bounds in Komlós' setting. Up to polynomial factors in the input size, the optimization problem can be solved in time $2^{O(k)}Δ^2$, and the corresponding feasibility problem in time $2^{O(k)}Δ$. We would like to point out that the original version of this work, published in ESA 2026 proceedings, contains a slightly weaker result. More precisely, the earlier bounds were $O(κ_k)^{2k} Δ^2$ for optimization and $O(κ_k)^{k} Δ$ for feasibility, where $κ_k$ denotes the maximum $\ell_\infty$-discrepancy in Komlós' setting. However, following the advice of one of the anonymous reviewers -- to whom we are very grateful -- we replaced the $\ell_\infty$-discrepancy with the $\ell_2$-discrepancy in the analysis. This is what allowed us to replace dependencies of the form $κ_k^{O(k)}$ with $2^{O(k)}$.