Mixed-type data clustering: a full factorial benchmarking study on distance-based clustering methods

This repository includes the code used for the project "Mixed-type data clustering: a full factorial benchmarking study on distance-based clustering methods", submitted as the first mini-project for the StatML CDT.

• Author: Efthymios Costa
• Supervisors:
  • Dr. Ioanna Papatsouma (Imperial College London) (main supervisor)
  • Professor Alastair Young (Imperial College London) (co-supervisor)

Description of files

• fullfactorialexperiment.R: R code for the full factorial design on data sets with non-spherical clusters.
• fullfactorialexperimentsph.R: R code for the full factorial design on data sets with spherical clusters.