Multivariate data analysis finds patterns in high dimensionality data by examining the interrelationships of the multivariate data set. Multivariate techniques can be employed to analyse complex data sets involving longitudinal data, spatial data (e.g. structural image of retina) and functional data (e.g. function of eye via visual acuity test). The most commonly applied multivariate techniques are discriminant function analysis, cluster analysis, principal components analysis, factor analysis, multidimensional scaling and MANOVA amongst others.

Some of our current work and applications include:

  • Development of dynamic discriminant statistical approaches for clinical prediction.
  • Development of dimension reduction techniques for selecting predictive features without losing important information that can affect classification accuracy.
  • Application of multivariate modelling techniques to identify clinical and biological predictors of response to a cell-based therapeutic approach aimed to treat Graft versus Host Disease.
  • We apply the methodology we develop to the clinical areas of diabetes, cancer research, brain imaging, stem cell research, ophthalmology (e.g., classification of patients with diabetes into diabetic retinopathy stages for early diagnosis).
  • We contribute to the development and implementation of a major enhancement to the existing delivery of screening for sight threatening diabetic retinopathy (STDR) by introducing an individualised approach based on patient-specific risk.

data modelling graph