Coherent Local Explanations for Mathematical Optimization
Daan Otto, Jannis Kurtz, Ilker Birbil · 2025
Read Coherent Local Explanations for Mathematical Optimization by Daan Otto, Jannis Kurtz, Ilker Birbil (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-04840.txt.
Description
Daan Otto, Jannis Kurtz, and Ilker Birbil explain optimization decisions in a 2025 paper. Managers who rely on algorithms for routes or schedules increasingly ask why a solution came out as it did. Explanation tools borrowed from machine learning ignore the structure of the underlying optimization problem, so their answers can mislead.
Their method, CLEMO, explains several parts of a solution at once, the objective value and the decision variables, in ways that stay consistent with the model's own constraints. A sampling procedure probes how the solution shifts as inputs wobble, working for both exact solvers and fast heuristics. Tests on shortest path, knapsack, and vehicle routing problems show the explanations behaving sensibly. The authors aim at trust: an explanation that respects the model is one a planner can act on. The work joins a growing push for transparent operations research.
Catalog record
| CREATOR | Daan Otto, Jannis Kurtz, Ilker Birbil |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-04840