Karen Kopciuk

My research interests include survival and multi-state models for estimating genetic and disease risks in families who carry a known mutated gene or putative genetic factors. Feature selection methods such as the group bridge estimator for high dimensional data including survival and other outcomes is another active research area. Lastly, projection to latent structures methods for metabolomics biomarkers measured over time are being developed to identify patterns of change. A current research project is evaluating factors associated with stage of cancer diagnosis to identify associated patterns of health behaviours using the Alberta Tomorrow Project data. Methods for ordinal scale response data and feature selection for complex, high dimensional health predictors factors are being created to address data-specific issues.