Publications
Peer-reviewed publications and preprints
Journal Articles
Some density-based silhouette diagnostics for soft clustering algorithms (2024)
Shrikrishna Bhat K and Kiruthika C
Communications in Statistics: Case Studies, Data Analysis and Applications, 10(3–4), 221–238, 2024.
DOI: 10.1080/23737484.2024.2408534
One of the main objectives of cluster analysis is to determine the most effective clustering algorithm. With the wide variety of algorithms available, assessing which one performs better is important. The performance of different clustering methods is typically measured using the Adjusted Rand Index (ARI), which relies on knowledge of the original class labels. However, this study introduces flexible modified alternatives of density-based silhouette methods for evaluating cluster performance. These proposed Density-based silhouettes can be applied to any soft clustering algorithms and do not require the original class labels. Instead, they rely on posterior probabilities. In this study, eight different soft clustering algorithms were evaluated using real and simulated data sets. The goal is to compare their effectiveness and performance using existing and proposed measures based on silhouette and the ARI.
Preprints
Block Probabilistic Distance Clustering: A Unified Framework and Evaluation (2025)
Shrikrishna Bhat K and Kiruthika C
Preprint, Version 1
DOI: 10.21203/rs.3.rs-6973596/v1 ResearchSquare: rs-6973596/v1
Probabilistic Distance (PD) clustering is a flexible and widely studied method in cluster analysis, owing to its probabilistic framework that combines distance measures with cluster membership probabilities. Building on this approach, we propose a novel block clustering framework and algorithm. The proposed algorithm is validated using both non-parametric distances, such as Squared Euclidean and Squared Mahalanobis distances, and parametric probabilistic distances derived from Gaussian and location Scale t-distributions for continuous data. To evaluate the clustering performance of the proposed algorithms, we modified the existing Extended Silhouette Index and used it alongside the established Co-clustering Adjusted Rand Index for comparison. This comprehensive evaluation highlights the effectiveness of our framework in advancing block clustering methodologies.