The Computational Complexity of Avoiding Strict Saddle Points in Constrained Optimization
Andreas Kontogiannis, Ioannis Panageas, Vasilis Pollatos · 2026
Read The Computational Complexity of Avoiding Strict Saddle Points in Constrained Optimization by Andreas Kontogiannis, Ioannis Panageas, Vasilis Pollatos (2026) 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-2604-02285.txt.
Description
Kontogiannis, Panageas and Pollatos settle how hard it is to find second-order stationary points, the stable resting points sought in non-convex optimization, when constraints apply. They prove that computing an approximate such point under the tractable definition is PLS-complete, even inside the two-dimensional unit square. The result bars any efficient deterministic iterative method with a continuous update rule, unless PLS sits inside PPAD.
Catalog record
| CREATOR | Andreas Kontogiannis, Ioannis Panageas, Vasilis Pollatos |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 216676 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.02285