CRB-NCE: An adaptable cohesion rule-based approach to number of clusters estimation
Year: 2025
Type: article
Abstract: Accurate number of clusters estimation (NCE) is a central task in many clustering applications, particularly in connection with prototype-based k-centers methods like k-Means, which require the number of clusters k to be specified in advance. This paper presents CRB-NCE, a general cluster cohesion rule-based framework for NCE that integrates three main innovations: (i) the introduction of tail ratios to reliably identify decelerations in sequences of cohesion measures, (ii) a threshold-based rule system that supports accurate NCE, and (iii) an optimization-driven approach to learn these thresholds from synthetic datasets with controlled clustering complexity conditions. Two cohesion measures are considered: inertia (SSE) and a new, scale-invariant metric called the mean coverage index. In this work, CRB-NCE is mainly applied to derive general-purpose NCE methods, although it also enables producing specialized procedures with enhanced performance under specific conditions, such as particular clustering algorithms, high-dimensional data settings, and overlapping cluster structures. Extensive evaluation on Gaussian and shapes synthetic datasets and real-world benchmarks shows that CRB-NCE methods consistently outperform or match standard internal clustering validity indices (ICVIs) in both accuracy and robustness. These results suggest that CRB-NCE may enable the recognition of cluster cohesion-based patterns not fully exploited by ICVIs.
Source: Pattern Recognition
Authors: J. Tinguaro Rodríguez, Xabier Gonzalez-Garcia, Daniel Gómez, Humberto Bustince
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $2,590
Awards: G5711763844, G696262180, G839282573
OpenAlex ID: W7115012492