Our research focuses on cluster analysis, revealing hidden similarities among data points and creating meaningful groupings. Despite challenges posed by high-dimensional data, we address this by identifying cluster-specific relevant features. Using a center-based clustering approach, we simultaneously determine cluster centers, assign data points to clusters, and select the most pertinent features. To achieve this, we propose a powerful mixed integer mathematical model. Given the non-linear nature of the proposed model, we employ various mathematical techniques and have developed heuristic algorithms tailored to the unique characteristics of the problem. Through extensive experiments, we validate the effectiveness of our proposed models and algorithms.

See Full Text