Research

Publications

Preprints

  • Model-based clustering using copulas with applications. Kosmidis, I. and Karlis, D. (2014). ArXiv e-prints, arXiv:1404.4077.
  • Liquidity commonality does not imply liquidity resilience commonality: A functional characterisation for ultra-high frequency cross-sectional LOB data. Panayi, E., Peters, G. W., Kosmidis, I. (2014). ArXiv e-prints, arXiv:1406.5486.
  • Upside and downside risk exposures of currency carry trades via tail dependence. Ames, M., Peters, G. W., Bagnarosa, G. and Kosmidis, I. (2014). ArXiv e-prints, arXiv:1406.4322.
  • Supervised sampling for clustering large data sets. Kosmidis, I., Karlis, D. (2010). CRiSM working paper 10-10.
  • On iterative adjustment of responses for the reduction of bias in binary regression models. Kosmidis, I. (2009). CRiSM working paper 09-36.

PhD thesis

Selected presentations

Projects for prospective PhD students

Below is a list of titles that I am highly interested in and I would be keen to research collaborating with a PhD student. Please contact me if you are interested on any of those to discuss or even arrange for a meeting. My contact details can be found here.

Optimal estimation and inference

  • Inference for models with many nuisance parameters.
  • Improved estimation for generalized linear mixed effects models.
  • The bias of functional estimators.
  • Estimation and inference for generalized linear models from big data sets.

Mixture models and applications

  • Model-based image segmentation with covariates.
  • Automatic identification of cell populations from flow-cytometry data.

Applications in Sports

  • Models for predicting "performance" in Sports.

The above list is certainly not exhaustive; if you have something else in mind and you are interested in working with me I would be more than happy to discuss.