Tim Biziaev – Using prior-data conflict to tune regularized regression models

Tim Biziaev, a MSc student in Biostatistics at the University of Calgary, successfully defended his
thesis on September 11. Tim was co-supervised by Professors Thierry Chekouo and Karen Kopciuk.
His thesis was entitled ‘Using prior-data conflict to tune regularized regression models’ where he proposed configuring shrinkage priors using checks for prior-data conflict between the data and the prior placed on the parameters. The method was applied to the Bayesian LASSO and spike and-slab shrinkage priors and compared to other variable selection approaches in simulation and real data studies for linear and logistic regression models.