AUTOMATIC GROUPING ALGORITHM BASED ON DENSITY AND GRID

The Automatic Clustering Problem (ACP) consists of partitioning a dataset into groups that are internally cohesive and externally well separated, without prior knowledge of the number of clusters. This work aims to solve the ACP by proposing a hybrid approach based on the integration of techniques used in spatial statistics with three metaheuristics implemented within the Random Key Optimizer (RKO) framework. To evaluate the proposed approach, experiments were conducted on 50 benchmark datasets from the literature, with varying numbers of objects and attributes, using the Average Silhouette Index (ASI) as the evaluation metric. The results indicate competitive performance compared to DBSCAN, one of the main density-based algorithms in the literature, showing that the proposed approach constitutes an effective alternative.