English

Continuous Convexity Measures

Geometric Topology 2024-12-24 v2 Optimization and Control

Abstract

Methods for measuring convexity defects of compacts in R^n abound. However, none of the those measures seems to take into account continuity. Continuity in convexity measure is essential for optimization, stability analysis, global optimality, convergence analysis, and accurate modelling as it ensures robustness and facilitates the development of efficient algorithms for solving convex optimization problems. This paper revisits the axioms underlying convexity measures by enriching them with a continuity hypothesis in Hausdorff's sense. Having provided the concept's theoretical grounds we state a theorem underlining the necessity of restricting ourselves to non-point compacts. We then construct a continuous convexity measure and compare it to existing measures. Importante note : This work is not a research article. It is an undergraduate project undertaken as part of a computer science course at \'Ecole normale sup\'erieure. It should therefore not be considered as a peer reviewed research paper.

Keywords

Cite

@article{arxiv.2306.02041,
  title  = {Continuous Convexity Measures},
  author = {Abel Douzal and Ferdinand Jacobé de Naurois},
  journal= {arXiv preprint arXiv:2306.02041},
  year   = {2024}
}

Comments

Importante note : This work is not a research article. It is an undergraduate project undertaken as part of a computer science course at \'Ecole normale sup\'erieure. It should therefore not be considered as a peer reviewed research paper

R2 v1 2026-06-28T10:55:22.000Z