Answers are grounded in this paper and cite their sources.
Algorithm NWGB-DPC enhances clustering performance in various densities, integrating natural neighbor structure and granular-ball modeling.
To address the limitations of Density Peaks Clustering (DPC) and traditional two-way clustering, we propose NWGB-DPC, an adaptive algorithm that integrates natural neighbor structure, granule ball modeling, and a three-way decision framework. NWGB-DPC adaptively estimates local density and distance based on natural neighbors, eliminating reliance on predefined cutoff distances and enabling robust detection of cluster centers in both dense and sparse regions. By combining Evidence Theory and Natural Neighbor information, it constructs a three-way decision framework that explicitly partitions objects into core and boundary regions. Extensive experiments on UCI benchmark datasets demonstrate that NWGB-DPC achieves superior self-adaptability and delivers state-of-the-art clustering performance.
No takes yet. Share an insight, caveat, or question.
Yan et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: