This repository includes the code used for the project "Investigating robust partitional clustering methods", submitted as the second 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)
robustclustsim.R: R code for the full factorial design on artificial data sets with well-separated components generated from the Gaussian and the t distribution. The clustering methods implemented are Multivariate Contaminated Normal Mixture (MCNM) Model, the Multivariate Normal Mixture (MNM) Model and the Multivariate t Mixture (MtM) Model on the Gaussian mixtures, while for the t mixtures, the MNM Model has not been implemented (see project for more details).