(Near)-Optimal Algorithms for Sparse Separable Convex Integer Programs
Abstract
We study the general integer programming (IP) problem of optimizing a separable convex function over the integer points of a polytope: . The number of variables is a variable part of the input, and we consider the regime where the constraint matrix has small coefficients and small primal or dual treedepth or , respectively. Equivalently, we consider block-structured matrices, in particular -fold, tree-fold, -stage and multi-stage matrices. We ask about the possibility of near-linear time algorithms in the general case of (non-linear) separable convex functions. The techniques of previous works for the linear case are inherently limited to it; in fact, no strongly-polynomial algorithm may exist due to a simple unconditional information-theoretic lower bound of , where are the vectors of lower and upper bounds. Our first result is that with parameters and , this lower bound can be matched (up to dependency on the parameters). Second, with parameters and , the situation is more involved, and we design an algorithm with time complexity where is some computable function. We conjecture that a stronger lower bound is possible in this regime, and our algorithm is in fact optimal. Our algorithms combine ideas from scaling, proximity, and sensitivity of integer programs, together with a new dynamic data structure.
Cite
@article{arxiv.2505.22212,
title = {(Near)-Optimal Algorithms for Sparse Separable Convex Integer Programs},
author = {Christoph Hunkenschröder and Martin Koutecký and Asaf Levin and Tung Anh Vu},
journal= {arXiv preprint arXiv:2505.22212},
year = {2025}
}
Comments
28 pages, will appear at IPCO 2025