Research

Research interests and focus areas

My doctoral research focuses on Block Probabilistic Distance Clustering (BPDQ) and its diagnostic measures — work that has resulted in peer-reviewed publications, R packages on CRAN, and invited conference presentations.

Research Interests

Cluster Analysis — Theory and methods for partitioning data into meaningful groups
Block Clustering (Co-clustering) — Simultaneous clustering of rows and columns of data matrices
Probabilistic Distance Clustering — Distance-based approaches with probabilistic foundations
Cluster Diagnostics — Silhouette-based and density-based measures for evaluating clustering quality
Statistical Computing — Development of R packages for reproducible research
Survey Analytics — Design and analysis of large-scale socio-economic surveys
Applied Statistics — Regression modeling, multivariate analysis, and data-driven inference

PhD Research

Contributions to Block Clustering and its Diagnostic Measures

Details

Institution: Pondicherry University Supervisor: Dr. Kiruthika Period: February 2021 – December 2025 Thesis Submitted: 22 December 2025

Key Contributions

Density-Based Silhouette Diagnostics — Proposed novel silhouette diagnostics for evaluating soft clustering algorithms, extending the classical silhouette framework to accommodate fuzzy and probabilistic cluster memberships.
Block Probabilistic Distance Clustering — Developed a unified framework for block (co-)clustering using probabilistic distance methods, with theoretical foundations and extensive empirical validation.
R Packages — Built and published open-source software:
- Silhouette — CRAN package for proximity measure based diagnostics
- blockclusterPDQ — R package for block probabilistic distance clustering

Research Output

Type	Count
Peer-reviewed journal articles	1
Preprints	1
R packages (CRAN / GitHub)	2
Conference papers presented	3
Poster presentations	1

📖 See all publications →

View complete list of publications, preprints, and conference papers.