English

On Densest $k$-Subgraph Mining and Diagonal Loading: Optimization Landscape and Finite-Step Exact Convergence Analysis

Social and Information Networks 2026-02-03 v4 Data Structures and Algorithms

Abstract

The Densest kk-Subgraph (DkkS) is a fundamental combinatorial problem known for its theoretical hardness and breadth of applications. Recently, Lu et al. (AAAI 2025) introduced a penalty-based non-convex relaxation that achieves promising empirical performance; however, a rigorous theoretical understanding of its success remains unclear. In this work, we bridge this gap by providing a comprehensive theoretical analysis. We first establish the tightness of the relaxation, ensuring that the global maximum values of the original combinatorial problem and the relaxed problem coincide. Then we reveal the benign geometry of the optimization landscape by proving a strict dichotomy of stationary points: all integral stationary points are local maximizers, whereas all non-integral stationary points are strict saddles with explicit positive curvature. We propose a saddle-escaping Frank--Wolfe algorithm and prove that it achieves exact convergence to an integral local maximizer in a finite number of steps.

Keywords

Cite

@article{arxiv.2410.07388,
  title  = {On Densest $k$-Subgraph Mining and Diagonal Loading: Optimization Landscape and Finite-Step Exact Convergence Analysis},
  author = {Qiheng Lu and Nicholas D. Sidiropoulos and Aritra Konar},
  journal= {arXiv preprint arXiv:2410.07388},
  year   = {2026}
}
R2 v1 2026-06-28T19:15:15.900Z