Algorithms for Standard-form ILP Problems via Komlós' Discrepancy Setting
For researchers in integer programming and parameterized complexity, this provides improved theoretical algorithms for ILP with small k and Δ, though the practical impact is limited due to the exponential dependence on k.
The paper presents refined FPT algorithms for standard-form ILP problems parameterized by the number of rows k and the maximum minor Δ, achieving running times that depend on the discrepancy bound κ_k. Using current bounds, the optimization runs in O(log k)^{k/2(1+o(1))}Δ^2 and feasibility in O(log k)^{k/4(1+o(1))}Δ; under the Komlós conjecture, these become 2^{O(k)}.
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 discrepancy bounds in Komlós' setting. Let $κ_k$ denote the maximum discrepancy over all matrices with $k$ columns whose columns have Euclidean norm at most $1$. Up to polynomial factors in the input size, the optimization problem can be solved in time $O(κ_k)^{2k}Δ^2$, and the corresponding feasibility problem in time $O(κ_k)^kΔ$. Using the best currently known bound $κ_k=\widetilde O(\log^{1/4}k)$, this yields running times $O(\log k)^{\frac{k}{2}(1+o(1))}Δ^2$ and $O(\log k)^{\frac{k}{4}(1+o(1))}Δ$, respectively. Under the Komlós conjecture, the dependence on $k$ in both running times reduces to $2^{O(k)}$.