Dr Keefe Murphy

Mathematics and Statistics

Lecturer

Logic House
130
(01) 708 4641

Biography

I am a statistician with primary research interests in developing methodologies for both supervised and unsupervised classification tasks involving complex, high-dimensional, mixed-type data, often in the presence of outliers. Publications to date include work on Bayesian nonparametric mixtures of factor analysers, covariate-dependent parsimonious model-based clustering, and risk-stratification for multi-omic prostate cancer data. I am currently working on applications of novel distance-based clustering approaches for longitudinal categorical sequences. I take a keen interest in the computational implementation of statistical models and have authored the R packages IMIFA, MoEClust, and MEDseq, all of which are available on CRAN.

Research Interests

I have a broad range of research interests in the fields of Statistics, Data Science, and Machine Learning including, but not limited to; mixture/latent-variable models, model-based clustering and classification, Bayesian statistics, shrinkage priors, applications to high-dimensional, mixed-type data, computational statistics, dimension reduction, sequence analysis, missing data mechanisms, tree-based methods, non-parametric regression, model selection, model aggregation, and data visualisation. Most of the research projects I have worked on involved a non-trivial computation aspect and I take a keen interest in distributing software implementations of my novel methodologies.

Peer Reviewed Journals

  Year Publication
2020 'Infinite mixtures of infinite factor analysers'
Murphy K.;Viroli C.;Gormley I.C. (2020) 'Infinite mixtures of infinite factor analysers'. Bayesian Analysis, 15 (3):937-963 [DOI] [Details]
2020 'Gaussian parsimonious clustering models with covariates and a noise component'
Murphy K.;Murphy T.B. (2020) 'Gaussian parsimonious clustering models with covariates and a noise component'. Advances in Data Analysis and Classification, 14 (2):293-325 [DOI] [Details]
2018 'Integrating biomarkers across omic platforms: an approach to improve stratification of patients with indolent and aggressive prostate cancer'
Murphy K.;Murphy T.B.;Boyce S.;Flynn L.;Gilgunn S.;O'Rourke C.J.;Rooney C.;Stöckmann H.;Walsh A.L.;Finn S.;O'Kennedy R.J.;O'Leary J.;Pennington S.R.;Perry A.S.;Rudd P.M.;Saldova R.;Sheils O;Shields D.C.;Watson R.W.; (2018) 'Integrating biomarkers across omic platforms: an approach to improve stratification of patients with indolent and aggressive prostate cancer'. Molecular Oncology, 12 (9) [DOI] [Details]

Honours and Awards

  Year Title Awarding Body
2020 Distinguished Dissertation Award The Classification Society

Education

  Year Institution Qualification Subject
2019 University College Dublin Doctor of Philosophy Statistics
2014 University College Dublin Master of Science Statistics
2012 University of Limerick Bachelor of Science Economics & Mathematics

Reviews

  Journal Role
Statistical Analysis And Data Mining Reviewer
Statistics And Computing Reviewer
Annals Of Applied Statistics Reviewer
Journal Of Computational And Graphical Statistics Reviewer
Computational Statistics And Data Analysis Reviewer

Teaching Interests

I am currently lecturing modules on Linear Models, Statistical Computing with R, and Nonparametric Statistics. Other courses I have previously taught include Statistical Machine Learning, Stochastic Processes, and Design & Analysis of Experiments.