The final project for the course was focused on the comparison between "unusual" clustering methods, namely model-based, grid-based and density-based. For each family the most representative algorithm was chosen. The project was made by three people, the part in which I was most involved was the grid-based section and the CLIQUE algorithm, both in the theoretical and test parts.
The repository contains the following files:
Report
: pdf file containing the report of the project, with a theoretical overview of the algorithm and a test part on both synthetic and real data.CLIQUE
: R file containing the tests performed
Links to the datasets used can be found in the footnotes of the report.